_Holy Mackerel!_ When I sent you that book for a light summer reading a few years ago, I did not expect _this_ to happen! Super excited and pleased and so much looking forward to it all!
Im glad you sent it to him! Very excited for the upcoming build series. I also agree, the motor is exceptionally pleasing to the eye. Very excited to hear how it will sound. Cheers from California!
Two things that will likely solve your core issues: 1. Increase the draft angle of the ribs connecting the pin boss to the bottom of the piston face. For small areas like that with lots of surface area trying to lock the material in there, extra draft is recommended. 2. Those problem zones are in areas that aren’t within the direct path of diffusion between your existing gas ports and their exit, so they don’t get proper gas flow. Add gas injection ports that feed directly into the trouble zone for proper hardening. Good luck! Excited to follow the project!
Yeah, I rechecked the model, and had only given it 1.5° of draft angle there. My bad. I have three gas injection ports close by, but will move them on the next print to direct the CO2 better to that problem zone.
I thought the exact same thing, and if there is a hole for gas injection right there is would also work as a 'vent' to stop any vacuum effect which may be grabbing at the core.
Thanks Paul. I have got to V5 of the core box and it is working pretty well now. I have to finish my new oven ths week, and then into the first castings for the Redrup radial.
You mentioned it right at the end of the video, but I, too, was wondering the narrow end of the mold had enough opportunities for the CO2 to get to the sand.
Sand is very porous, and the three extra hole I put in to push in gas came within a couple of cm of that point, but I will move those channels directly to them.
Hey Mark, since you're 3d printing the molds you could consider making sub-mold parts that come out with the core and are removed as another step. The narrow area you've been struggling with could be made of two or more small parts and a sleeve that you can remove more gently after removing the main body. Also you could add ejector pins too. If incomplete setting of the silicate keeps happening you could bathe the mold in CO2 in a closed container since it's heavier than air. I don't know if dry ice would improve things at all, but that could be a good source of cold dry CO2. I've never done any of those, but I just wanted to toss ideas your way. Good luck and great idea for a video series.
Thanks for your ideas Jonathon. I rechecked my CAD model and had only given the pin boss and rib 1.5° of draft angle, which is no where near enough. Need to remodel it.
@@RotarySMP What he said. Instead of a 2-part mold, you could probably make a 4-part mold (where the pin sleeve and the web supporting it are two parts that get removed before you break open the bigger mold); you'd keep the pin going through to preserve alignment.
I’m sure your sister is going to be very proud of you. I too suffer from the tool cubed syndrome. When you talked about doing the piston first I was reminded of the opening scenes in the Worlds Fastest Indian. I’m looking forward to the series.
Heck yeah, I’m looking forward to more RadialSMP! Also: you should check out Project Hercules. He’s building a chopper around a single cylinder from a Bristol Hercules engine. I can’t quite put my finger on why, but it seems like it might be something you’d be interested in 😂.
Hi Jeremy, thanks for the tip, YT had never offered me that channel before. Subbed.The Bristol Hercules is going to be a big job with the sleeve valve. Fitting in the valve actuation will be a big challenge. Do you also watch Alan Millyards channel? His is the pioneer of such projects.
i love those three things! lets go.... re the sand cores, I think you just need better draft angle on the non filleted areas. Also i used to make a slightly strong mix too and CO2 vents direct to the troubled areas. fun project, looking forward to more!
@@aaronhammond7297 The model was drafted, but you get into a corner of trading draft angle for wall thickness for fillet radius in these small features.
Epoxy instead of sodium silicate works well for thin and complex cores. You don't need to worry about the CO2 penetrating, and it is really strong. Printing in some flexible joints on your big core mould makes it easier to get the core out. This is going to be an epic series. Regards, Preso
Hi Mark, I rechecked the model, and screwed up. I'd only given the pin boss and rib 1.5° of draft which is insufficient for such and enclosed feature. I'll redo the model and reprint it. I may need epoxy cores of moulds for the crank case, but it is nice to be able to iterate so fast with silicate sand.
Ah man can't wait to see this project come together! Have been thinking about starting on my own gas engine projects but for now I'll stick with little sterling engines.
When you just want to take a break from your classes and your favourite machining channel now also has a "required reading to follow the course contents"... I love it. Time to procrastinate
And being a triple cylinder there are going to be at least 3 of lots of parts, just enough to make cnc worth it? And multiple of bolts and the like, worth dialing them in. But not dozens to get boring. Great project for your shop and skills.
As soon as you said refining the design slightly, I thought it'll end up up a 6 cylinder radial engine. But aeroplane engines and motorcycles, cant get any better. Thanks from the UK.
I'm glad you are not having any more crash videos, I do like your content but I couldn't bring myself to watch that! As a machinist I've never had to do any casting myself, so really I'm looking forward to seeing your progress. So far it looks to be pretty frustrating!
Hi Andrew, thanks. It is probably good practice to stand on the little pistons, as I am sure the crankcase with integral cylinders is going to be a real learning experience for my little foundry :)
Past employ as GE apparatus service dividion and years later J Deere maint (shop worker), I am a new subscriber, WA state, US. I own a S Bend 9 of 1935 vintage. I am delighted to watch the progress to this curious design and sort of astonished as to the strict build to extracted plan. I would change up anything as found useful, you see. I repair various Honda superbike (1986) CB1100 as carburetor type. A 98 Jeep cherokee limited *(grey leather) 53k original miles and find myself cheering for your project. As a young teen, I chanced to visit a site with steam and various machines of 1910 to 1939 vintage (Mogul, other fuel) thus appreciate vintage iron. Semi-retired, I have access to a huge scrap yard with rather the time and inclination to fitter, weldment projects. Thanks for having me along. M. For the hobby or professional I recommend Max Grant, Swan Valley (AU) Shop's channel. Also, IC Weld. Looking forward to progress.
Yay! MOTO content! I'll have to agree with your Sis, tools for tools are nice but in my shop everything has to justify it's existence. I'm a moto builder who skilled up in these areas to make parts. Salivating for this series.
I would suggest putting your sodium silicate cores still in the molds into an air-tight container, pull a vacuum, and then flood it with CO2 and let it stand over night, this will allow more of the sodium silicate to harden, making de-molding,and cleaning up the pattern cores easier.
Trust you to pick an easy one for the first project! Really looking forward to this one!! You could try a stronger mix of sand with more sodium silicate or can you get some holes right into the problem area to make sure the gas is getting to it? Redrup is one of my heroes! Phil Ps BSF into aluminium is not reliable. Royal Enfield among others used them and I have spent a fair time helicoiling them!!
Hi Phil, thanks. I did put those little gas channels in, but I think I need to move then down by about a half inch to get the gas where it needs to go. I dont think there is much aluminium in this engine other than the heads and pistons. Need to take a close look to see if the timing gear covers / main bearing carriers are iron or alluminium.
@ I wanna build an engine based on 1930s Husqvarna racing single 350cc with overhead valves. But base it around a Vw 1200 cylinder. Will probably try to use some existing parts, like cylinder, valve, cams
BSF is a torture choice, the best of luck :D. Have the plans and raw castings for a BSA B50 V twin that I've not started yet so kudos for doing the first step and getting a project going. Looking forward to watching you carry on this project.
This will be an interesting project and likely lead to a huge amount of content. I am already rubbing my hands at the thought of all of the side-projects that will come along with the series. I wish you well🤝
Dang, I was hoping for an axial compression jet engine build. Maybe next project. Looking forward to this build. And found you by the video for the doall bandsaw gear box fixes, I have the same saw. Cheers and really enjoy your videos.
I helped make parts for aerospace companies such as Boeing, Korean Air, Liebherr, the Military etc. I’m into aerospace history just a tad. Can’t wait for this project of yours. After this build you should consider making a copy of the Burt Munro 600cc modified Indian engine that he built for his Indian motorcycle, which still holds the record for the fastest Indian motorcycle. I lived in Utah when he set the record and I was about 10 miles away in a small border town called Wendover. I would have been thrilled to see the event in person, but I was just a very young kid in 1967. 😊
Yeah, that would be a cool project, but I would imagine that those engines are now rare as hens teeth, and it would be tough to fine either drawings or someone willing to let theirs be disassembled and measured.
Good day to you sir. I would really appreciate you replicating this motor. You have demonstrated your skill clearly in your previous videos. Your replication will help us understand how a man would approach this over 110 years ago. I couldn't think of a better man to undertake this project.
I basically came to the same decision you did last year after spending nearly 3 years just building my machines and tools to use on them, that it was about time I started using them to make things with. I had been watching a model steam engine build by Mark Presling and I decided I would tackle one of those, but he made one from a kit and I decided I wanted to build one from scratch with my own castings. However, my casting were only Aluminium and not Cast Iron like the kits, but I had no intention of ever running it on steam, so that wasn't an issue. After I started on that, one of my long-time viewers commented that in his opinion making model engines was the first sign of a dying channel, I just shrugged that off at first. During that build another long-time viewer suggested that I should tackle an IC engine next up, and I ran with that idea and have spent the last 2 months building a Webster Engine which I've nearly completed. However, I have been watching my view count steadily decline over both builds, and I'm now starting to wonder if the Dying Channel comment was kind of right, I sincerely hope that doesn't happen to you. On another note, to save a lot of time and hassle you should consider Lost Foam Casting, in which you can substitute the foam for 3D printed items, but they need to have hardly any infill so it will burn away easily, I have made a lot of lost foam castings and once you get the hang of it, it's a real time saver, and they come out pretty well. I just this afternoon ordered a 3D printer to give it a whirl as a week ago I bit the bullet and downloaded FreeCAD 1.0 and got stuck into learning that. I'm personally looking forward to watching this build. Sorry about the long comment.
Interesting feedback. Thanks. I was once going to build an Edwards Radial 5 model, but decided that I dont really want to do model engineering. Maybe I am kidding myself, that this is not a model build but an actual replica of a motorbike engine, so I would also have the possibility to continue on and build the frame and motorbike for it. I sort of considered lost foam casting, but this engine is still small enough that I should be able to cast the crank case with cylinders conventionally. I will start making at least two sets of everything, and even more with pistons. I do like what Kelly Cofield is doing with lost foam though. I want the multiples of parts to force me to set up CNC machine jobs properly, and not just dick around with one offs :)
@@alanclarke4646 It's the one thing that had been stopping me from buying one, plus a complete lack of somewhere to store it, so it looks like it will have to live on the dining table when I'm not using it, which won't please she who must be obeyed.
Mark, it looks like you have no draft on your mould. There should be no vertical faces when you look down on the two mould halves. Did you add a shrinkage allowance to your mould? 😊
You got it. I only gave that pin boss and rib 1.5° which is not enough for such an enclosed feature. I'll redo it. Yes on the shrink allowance. Nice thing about 3D printing. You just design the model nominal, and then scale the STL model 2% in the slicer.
I highly recommend to make 3D prints of the most complex parts to test that the interaction between components to make sure that the engine will turn over without interference. I’m sure that there are parts where the drawings are ambiguous, and holding a 3D model in your hands will help a lot.
Great project, thanks for doing this, I look forward to watching the whole series. Btw if you want to remove the facets on your 3d printed FreeCAD parts, open up the settings, go to the "Part/Part Design" section, then "Shape view" under that, and reduce the "Maximum deviation depending on the model bounding box" to maybe 0.02%.
Thanks for the tip, I did wonder why they were so faceted, but figured the sand would not care :) I just went in and adjusted it. Thanks. Actually, I had gone looking for this setting earlier on, but had no idea what it was called and didn't fine it.
Hi Mark., Good luck with you journey. As it was already noted earlier these zones are not getting enough CO2 to harden. you can drill additional holes that will allow to harden this better OR you can use epoxy glue instead of calcium siliicate - epoxy will cure itself - no gas needed and it will cure evenly on all core.
What an incredible project! 3 cylinder bikes are making a comeback, maybe you can help bring back radial bike engines too. 😁 Re the sand molding, looks like the strength of the sand is too low for the strength needed to hold those detail "ears". I recommend really increasing the draft angle on your cad model for any small details like those ears. You looked to have almost no draft angle at all and with such a weak material (crumbly sand) you need a heap of draft angle like 12 or 15 degrees. Another thing would be to increase the strength of the sand with more binding agent, making it less crumbly. And as a final suggestion, I'm not sure why you are making that tiny sand core rod for the piston pin? I think the pour will work better without that complexity, less chance of air entrapment or premature metal cooling if you just make the piston with solid metal in that area. After all, drilling the piston pin bore is one of the easiest jobs, personally I would prefer to drill solid metal there and not try to drill in a hole full of sand particles and I think it will just mold easier. But then maybe I'm just missing something and there is a good reason for the core?
Yeah, I had draft, but just went back in and checked my model and only have 1.5° there. Normally I use at least 3° on any internal surface, and as you point out, that little corners need more than that. I will revise the core model. I guess I was just following the book, but also wondered whether that pin core gains me anything. Normally it would reduce the risk of a shrinkage error from large concentrations of metal, but this is a pretty tiny piston.
@RotarySMP ooh yeah 1.5 degrees is the kiss of death lol. 5 or 6 is a minimum for a weak material in a non compliant (ie; solid) mold, and then for small details you need to increase far above the 5 or 6. Maybe you are overthinking the complexity needed in the piston core cavity? People make perfectly good pistons out of a slug of solid metal, just cutting out the conrod cavity with an endmill. I know you want to mold the piston, just saying maybe the core sand slug doesn't need to be that complex? Anyway thanks for the videos, the process is as rewarding as the success, whichnis why we do this kind of stuff I guess 🙂
Wax mold release on PLA 3d printed molds takes a few heavier layers. On composite parts I've done I've had to add a layer, wait a while for it to completely dry, and repeat at least 3 or 4 times.
I'm about to start down this path myself. I have a 1920's designed inline 4, water cooled engine that needs new oversize pistons, and the original pistons are cast iron. Not having a piston cam grinder to remake a similar size piston with proper skirt clearances, I am looking at either machining my replacement pistons from solid, or making the core molds as you're doing here. I've had similar success with 3-d printed core molds, and a lot of the issue I had was layer line direction. I watched some videos by Easy Composites here one youtube where they were making 3d printed molds for carbon fiber parts and I'm going to try to implement some of their suggestions and methods.
What a project 🎉 very looking forward to it........ for your sand mixture: glucose is a honey like liquid and would glue the sand together like epoxi.......sucrose is just crumbly and just makes your sand sweet.
Hi Christan, it is not about the cold sand, it is about weakening the sand once it has been backed by the hot metal. You want that core to collapse, without the sugar, it will be a bit concrete like and hard to remove.
gloucose is like honey, it will spread more even and i think it will caramelize and burn to carbon when you pour and remove all adhesion so it will cruble quite easy afterwards........ just my expirience from cooking 😅
Looking forward to seeing you progress through this project, very exciting! On the piston casting, could you rotate the orientation 90 degrees so that the wrist pin features are split half to each side for better sand plug removal?
I think you should prime and polish your mold and put a pin hole in each of those corners that break out. Additionally if your CO² dispensing shroud was slightly concave and covered the entire mouth of the mold it would help force the gas down into the corners. I havent done any casting since highschool, but i dont think there's any way that raw plastic from the printer is going to release cleanly.
Hi Jim, I am really surprised it releases sand as well as it does. With my old printed I had to fill and prime to remove the striations, but on the Bambu I am printing at 0.08mm layers and it has very little striation. The big issue turned out to be draft. I rechecked the model, and had only given the pin boss and rib 1.5°. I redesigned it and reprinted it, adding more gas holes, and at the same time printed a full size CO2 shroud (I was meaning to already). On a work trip today, but looking forward to trying the new mould. It would be nice if I can get away without painting and polishing it. Obviously if it still hangs up, I will have to.
Wow this looks like fun. I've found that UV cure epoxy is an excellent alternative as you can build up a structure gradually and exactly where it's needed layer by layer. A kind of manual resin printing if you will. Also, as it is the core, would making it in sections with a final glue up work? ie expand the whole gudgeon pin core to include web and fillets and then insert it into one larger hole in the piston wall.
Did you forget to add draft angles around the area of the 3d printed mold for the sand core since it kept refusing to release? didn't seem like you had any from the video
Thanks Brian. The sand is so porous, it should be easy enough to get all areas in that little core gassed. I probably just didn't hold the pipe there long enough.
Glad that you're apparently over whatever virus that you had. Judging from the beanie/hoodie/sweater (jumper) combo it looks like your basement is not any warmer than the 13C that you cited previously.
Ohhh wow!!! This project will be one to follow! BTW, have you looked into using epoxy for the core moulds? Mark Presling has used it successfully several times.
Epoxy core sand could make it worse, as it looks like a lack of draft. An epoxy sand core would not break, and could destroy the mould. I need to redo it with more draft.
A Megola rotary would be interesting...if not more challenging to build. I knew the British had several systems for threads which were largely incompatible but I didn't know they had one just for bicycles and motorcycles.
I remember rading about them. Looks like a suicide device to ride. Yeah, CEI (Cycle engineering institute) threads became BSC. I think it is something like all threads haveing 26TPI regardless of size, or something like that.
Nice one Vince. I have asked Bill if he has a link to a current publisher, as my copy was published by Diesel Publishing in Badminton, but I cant find anything about that. Once he gets back to me I will add another link.
Three random thoughts: 1. How much draft do you have on the wrist pin bosses? Maybe not enough, so that slight side to side twisting as you remove the core cracks it? 2. Maybe the sand isn't completely curing down in that part of the mold? Maybe a couple of CO2 injection ports right over the bosses, rather than just one at the bottom (or top, or whatever) of the mold? 3. If all else fails, print the mold out of something somewhat flexible, so you can peel it off the core?
1. Insufficiant. I rechecked the model, and only gave it 1.5° there. Need to redo it. 2. I have three CO2 injection ports , and will move them even closer where they need to be. 3. Good idea, a little less rigidity would also help.
I have three gas canals to that area, but can probably optimise their location. Once I get this core mould dialed in, it is pretty quick to cast a whole batch of pistons. It is no trivial to machine those voids. They are deep, narrow and have undercuts.
This looks like an excellent project that I am going to enjoy watching. Looking at the problems you are having with your cores, I would hazard a guess that the problem is a combination of not enough gas exposure in those areas of the core box and also that the sand might not be packed firmly enough...
Yeah, I can see that some of them I packed better than others. The silicate sand doesn't pack down the same way green sand does, but still needs more packing than I gave it.
I am pretty sure it is a shape problem Robert as I rechecked the model and only have 1.5° of draft angle there, although on some of the attempts I might have undergassed them,
I expect your CNC milling machine can be used to clean up a looser casting but they did manage a solution in 1920. It might be worth sending a DM to olfoundryman and seeing if he will answer.
@@robertfontaine3650 Oh Martin and I have been in close mail contact for the last few weeks, and once he sees this video (middle of the night in Oz), I am sure he will get back to me. I screwed up the draft. I checked the model and only had 1.5° around the pin boss and rib. I need to redo that.
Very nice project! Looking forward to see more 🙂 I don't have any knoweledge about sandcasting, but I would try to split the core form into 4 instead of 2, to reduce the stress when you pull it apart.
I hadn't given it enough draft angle. A new mould was printed yesterday. If I do need to split the core mould more, I will slice off the bottom and make it removable.
Would using a three part mould for the cores, with a separate bottom piece, be any better? My thinking is that the less surface your trying to separate at once might help prevent breakages. Just a thought.
I couldn’t click like fast enough. I’m excited for this. I like the getting started approach. I wonder if painting the prints with a high sheen paint will help with release. What draft angle are you using? For one of my latest patterns everything has a 2* taper. In fusion it’s handy that you can add draft to any extrusion.
I rechecked my model, and had only used 1.5° draft angle, which is too little for such an enclosed internal feature. I will redo it with much more. FreeCad also has that option.
Decades ago when I ran a printing press, I would polish the back cylinders with Johnson’s paste wax and then spray them with silicone spray. Perhaps that might work for you as well?
Omg, this is so me. I work in IT but hobby have constantly improving my not small workshop anymore. I had always squint when getting this question: "what have you actually build" 😂. Nothing that I can think of is the answear. Im also thinking about building small engineering myself from some time but didnt get to the actual start part yet 😅
I was trying something new. Normally I dont do talk to camera, but had nothing else visual to use. Hopefully the further videos will have more to show.
@ It was thanks. I noticed when I edited it that I was watching the screen just off the camera axis, rather than looking at the lens, and also that I screwed up some takes and looked at the camera which was off (I got out of sync of starting and stopping the two camera and audio recorder, and ended up with some clips the camera recorded between dialog. Duh! :) With a face made for radio, I prefer the videos to be about the project rather than showing me.
Free Cad fillets, and the pain. I'm always incrementing them in 0.2mm until it breaks, then back off a little. Failing that, create the mirror profile, and do a boolean cut
Not yet. I am pretty sure a little more grinding and polishing in that area and it will start working reliably. Add ejection help may be an option if needed.
Using real mold release is lava. Re: the filets - it's the reason why Parasolid and ACIS are the most popular modeling kernels, since they can accurately compute filets intersecting with each other. If there's no particular reason for filets, you could try chamfers, since they are geometrically simpler to calculate.
Hi Alexander. FreeCad seems to hit a limit when a fillet would touch another feature beyond a certain angle. It is probably helpful for me to do this with epoxy and grinding, as I dont really want to keep printing waste plastic core moulds which dont work.
@RotarySMP Yeah, it's a foundational problem with the modeling kernel. One of my colleagues is a beta tester for Plasticity, he also makes tutorials here for that software, and mid-development, the developer switched from some kernel to Parasolid, since it caused to many limitations in what the software could calculate. Including filets and chamfers. It seems like a solved problem, but in reality relies on kernels made 50 years ago, that to this day are getting commercially licensed, without much chance for open-source software being able to reproduce it.
@@graealex Freecad is certainly a victim of many of the weaknesses of open source software, but for the most part I can make it work for me, and it is improving.
@@RotarySMP It's going to be hard to overcome this fundamental flaw. The big CAD players' shopping sprees made sure of that. And fillets aren't just a cosmetic thing, they are integral since real-life tools seldomly create sharp corners.
@@graealex True, but this old engine is also not rocket science. If I can get close with Freecad fillet and a 3D print, it is not a deal breaker for me to go old school and add fillets of epoxy here and there.
I got the feeling, it breaks where you have low flow rates of CO2. Maybe try vent holes right in the middle of the stuck part or leave the whole thing in a nice warm CO2 bath over night. Anyway, wish you success!
this might be a good part to make with sacrificial plaster molds, the kind where you 3D print the core, embed it in a plaster mold and melt it out to make the negative space it does mean making 3 molds, one for each piston, but it might be easier
Oh, you mean investment casting. It is not really ideal for this kind of task, as you have to burn out the investment (takes about 12H). It is great for thinks which require accurate geometry but not engineering fits. In this case, yoou have to machine the OD, the ring grooves and pin bore so it is a waste to investment cast it. Also I dont need three, I need enough piston blanks to end up with three good pistons after inclusions, machining errors etc. :) Once I get this core mold dialed in, it will be much faster to sand cast a batch of pistons.
@RotarySMP yeah is probably much better for one-offs or complex geometry. with proper draft angles the pistons should be very doable with sand, although the housings themselves might be a different story when you get that far
@@terjestrm1796 I started modelling the crankcase with it's integrated cylinders. That is going to be a real challenge. I have also modelled the cylinder heads, and found them more difficult than expected. By the time I am ready to try casting the crankcase, I whill have a bit more experience in both casting and 3D printing core moulds and patterns.
question: was doing a lost PLA molt not attainable? (printing cylinder head positive with PLA, pack it into the sand, and bake out the PLA at 240C in the oven?
Investment casting is done with a plaster like material, and it takes 12hr to burn out each form. It could be used, but is a waste as investment casting gives very good accuracy, but not enough so you still need to turn the OD, ring grooves and pin bore. Once I have this core mould dialed in, it will be very fast to produce a batch of pistons. I need more than 3 as I'll have some with inclusions, and screw up the machining on some.
Book bought. *Really* looking forward to this series! Especially seeing your amazing machines in action. I think (being the expert I am (not)) increasing the draft angle of the gudgeon pin wall and its web (to the underside of the piston) would help more than increasing the fillet radius?? Also easier to do in FreeCAD ;-)
Hi Mark interesting new project. a suggestion from someone that knowns next to nothing about casting and cores. what if you prepacked each half of the mold first to ensure the sand around the bottom is truely compacted.
Hi Steve, this silicate sand doesn't compact down like green sand does. It sort of just moves around. I do need to compact it a little better, but am pretty sure that is not the main issue. I need to smooth out that epoxy fillet I added.
have you tried playing with drawers for your core mold? they are super expensive when injecting plastic, but you are demolding your cores by hand, you can pull on a piece of plastic for the same price. I would probably try to make the boss around the drill bit a drawer, pulling it towards outside should shear the bond between the mold and the core, and when opening the main mold the sand will be far less "trapped".
@@RotarySMP sorry, I used the wrong word, it looks like it's called "SIDE CORING OR SIDE DRAW PINS" it's like your drill bit, it's a separate part of the mold that you have to pull before opening the mold because it's not demouldable by straight pull. Basically I propose to 3D print your drill bit+its plastic shoulder in one piece (split in 2 at the parting line), and pull that first, then open the mold. My reasoning is that currently when you open the mold you have to shear 2 molded faces at once, by having a side core, you shear one face when pulling the side core, then the other face when opening the mold.
@@nraynaud Thanks. I have an email address on the channel main page. Would you be able to send me a sketch, as I am struggling to follow the description.
Perhaps some graphite to help with the corners. It was a common practice to use a water based spray with graphite in the silicate core dies. However they were usally not as detailed.
Thanks for the tip. I tried to find a spray bottle to do that with a while again, and found the nozzle would clog before it was useful. I need to find a solution for this before I get to the main crankcase iron casting, as that will be a bear and a graphite wash would probably really help.
@@RotarySMP We used to make simple air driven syphon guns. The crankcase may benefit with a core paint. During low pressure aluminum casting the naked cores would get worn and distribute sand in the casting. The core paint works well in high flow areas.
Brilliant! Looking forward to the rest of the build, I'd love to do something similar in the future Looking at those ears flaking off, do you think a few more gas holes might ensure they cure? Or would it make sense to cure halfway while the core is shallow?
I'm excited to see this project gain momentum! The piston ring retaining the gudgeon pin doesn't make a lot of sense, you have all the extra friction of the piston ring but it doesn't provide any sealing benefits as there is a direct path around it through the gudgeon pin hole. Is the lower ring supposed to be an oil control ring (doesn't appear to be)?
Could you get the mold in a vacuum chamber? Removing the air then adding CO2 would eliminate any soft spots. A pressure cooker could be modified as the vacuum chamber.
In plastic injection molding, the mold release we use contains PTFE. Not sure where you could obtain it where you are. Why are you building sand cores rather than using lost-PLA casting?
Thanks. Once I get the core mould dialed in, sand casting is way faster. Investment casting requires each mould to be burnt out for 12H, and the investment is expensive.
Oh this is gonna be so interesting, thank you Mark. The sand cores, when it has been used for the casting, can it be broken up and used again or does the Sodium Silicate (?) totally mess with the sand?
Typically you try to separate the core sand from the green sand during shake out. A bit of core sand will get into the green sand though, but I dont think it makes much problem. It will just get mulled into the green sand. I'll just dispose of the separated core sand. A 20kg bag of sand only cost a couple of € from te hardware store.
_Holy Mackerel!_
When I sent you that book for a light summer reading a few years ago, I did not expect _this_ to happen! Super excited and pleased and so much looking forward to it all!
Hi Hanrey, yeah that was a good read and has been worming into my conscience ever since :)
Im glad you sent it to him! Very excited for the upcoming build series. I also agree, the motor is exceptionally pleasing to the eye. Very excited to hear how it will sound. Cheers from California!
@@norcalmrbThanks for watching and commenting.
He who has the most tools at the end of life is the winner.
Yeah, my sister doesn't see it like that :)
Challenge accepted 😁
Pretty much !
@@RotarySMPas much ax i enjoy your content, i do have to side with your sister there 😂❤
greetings from vienna!
@@DavidSchmitt Stop being a girl 😆
Two things that will likely solve your core issues:
1. Increase the draft angle of the ribs connecting the pin boss to the bottom of the piston face.
For small areas like that with lots of surface area trying to lock the material in there, extra draft is recommended.
2. Those problem zones are in areas that aren’t within the direct path of diffusion between your existing gas ports and their exit, so they don’t get proper gas flow.
Add gas injection ports that feed directly into the trouble zone for proper hardening.
Good luck! Excited to follow the project!
Yeah, I rechecked the model, and had only given it 1.5° of draft angle there. My bad.
I have three gas injection ports close by, but will move them on the next print to direct the CO2 better to that problem zone.
I thought the exact same thing, and if there is a hole for gas injection right there is would also work as a 'vent' to stop any vacuum effect which may be grabbing at the core.
@@hugobiddlecombe504 That sand is pretty porous, so I doubt there is much vacuum effect there. The new core mould is on the printer now.
"What have you produced?" For the 30+k of us subscribed, joy, learning, encouragement and support.
Thank you.
Thank you very much for your kind words.
A fanbase
@@alanclarke4646 Thanks Alan.
I agree, providing education and entertainment is a worthy accomplishment.
@@jrmintz1 Thanks.
fellow lifelong two-wheel enthusiast and nearly lifelong two-wheel builder here saying heck yeah
Thanks Paul. I have got to V5 of the core box and it is working pretty well now. I have to finish my new oven ths week, and then into the first castings for the Redrup radial.
Very cool! Gonna be awesome...and to hear an old 3 cylinder radial run will be awesome!
Keep em coming!!!!
Thanks Hayden. I hope so.
You mentioned it right at the end of the video, but I, too, was wondering the narrow end of the mold had enough opportunities for the CO2 to get to the sand.
Sand is very porous, and the three extra hole I put in to push in gas came within a couple of cm of that point, but I will move those channels directly to them.
Hey Mark, since you're 3d printing the molds you could consider making sub-mold parts that come out with the core and are removed as another step. The narrow area you've been struggling with could be made of two or more small parts and a sleeve that you can remove more gently after removing the main body.
Also you could add ejector pins too.
If incomplete setting of the silicate keeps happening you could bathe the mold in CO2 in a closed container since it's heavier than air. I don't know if dry ice would improve things at all, but that could be a good source of cold dry CO2.
I've never done any of those, but I just wanted to toss ideas your way. Good luck and great idea for a video series.
Thanks for your ideas Jonathon. I rechecked my CAD model and had only given the pin boss and rib 1.5° of draft angle, which is no where near enough. Need to remodel it.
@@RotarySMP What he said. Instead of a 2-part mold, you could probably make a 4-part mold (where the pin sleeve and the web supporting it are two parts that get removed before you break open the bigger mold); you'd keep the pin going through to preserve alignment.
I’m sure your sister is going to be very proud of you. I too suffer from the tool cubed syndrome. When you talked about doing the piston first I was reminded of the opening scenes in the Worlds Fastest Indian. I’m looking forward to the series.
Thank Jim. Jury is out on whether my sister would be proud, or even notice. :)
Heck yeah, I’m looking forward to more RadialSMP!
Also: you should check out Project Hercules. He’s building a chopper around a single cylinder from a Bristol Hercules engine. I can’t quite put my finger on why, but it seems like it might be something you’d be interested in 😂.
Hi Jeremy, thanks for the tip, YT had never offered me that channel before. Subbed.The Bristol Hercules is going to be a big job with the sleeve valve. Fitting in the valve actuation will be a big challenge.
Do you also watch Alan Millyards channel? His is the pioneer of such projects.
i love those three things! lets go.... re the sand cores, I think you just need better draft angle on the non filleted areas. Also i used to make a slightly strong mix too and CO2 vents direct to the troubled areas. fun project, looking forward to more!
Thanks Julian.
Yeah the fillets probably aren't so much of a problem, just the lack of draft.
@@aaronhammond7297 The model was drafted, but you get into a corner of trading draft angle for wall thickness for fillet radius in these small features.
Maybe splitting the halves the other way would help with the problem areas? Just going on a limb here...
@@michaelguzzi1 Splitting through the center of the pin boss creates a huge undercut under it.
Epoxy instead of sodium silicate works well for thin and complex cores. You don't need to worry about the CO2 penetrating, and it is really strong. Printing in some flexible joints on your big core mould makes it easier to get the core out. This is going to be an epic series.
Regards, Preso
Hi Mark, I rechecked the model, and screwed up. I'd only given the pin boss and rib 1.5° of draft which is insufficient for such and enclosed feature. I'll redo the model and reprint it. I may need epoxy cores of moulds for the crank case, but it is nice to be able to iterate so fast with silicate sand.
Ah man can't wait to see this project come together! Have been thinking about starting on my own gas engine projects but for now I'll stick with little sterling engines.
Cool. Thanks for your support.
This is going to be a great adventure!
Hi Jim, I hope so as well.
When you just want to take a break from your classes and your favourite machining channel now also has a "required reading to follow the course contents"...
I love it. Time to procrastinate
Thanks, I hope you enjoy the series going forward :)
Looking forward for this project Mark. Its going to be great watching. Cheers
Thanks Steven.
And being a triple cylinder there are going to be at least 3 of lots of parts, just enough to make cnc worth it? And multiple of bolts and the like, worth dialing them in. But not dozens to get boring. Great project for your shop and skills.
@@stevensmart8868 I am planning to cast about 10 pistons, as I figure there will be inclusions and machining errors to whittle down that number.
Wow, that’s some technical foundry work, that is. I’m going to enjoy the ride. 👍
Hi Doric. I already started on the crankcase model, and can see that is going to really push the limits of my foundry work.
OK, ready for the next episode please.
Gonna be a fun ride, thank you.
John 🇨🇦
Will be a couple of weeks till the next episode as I really need to finish my new oven first, but then look forward to getting into those castings.
"What's the workshop for?"
My sanity!
I'm looking forward to this lots :-)
That is an excellent answer. :)
Oooo, looking forward to watching this later. Triples, sixes and twelves - all inherently balanced if configured properly 😁. Smooooth.
It is a nice little design, with the Knife and fork (and spoon?) conrod big ends, so you dont have the imbalance caused by a Master/slave rod design.
As soon as you said refining the design slightly, I thought it'll end up up a 6 cylinder radial engine. But aeroplane engines and motorcycles, cant get any better. Thanks from the UK.
Good call Dave, d you know how easy it is to got from 3 cylinders to 11 in CAD :)
I'm glad you are not having any more crash videos, I do like your content but I couldn't bring myself to watch that! As a machinist I've never had to do any casting myself, so really I'm looking forward to seeing your progress. So far it looks to be pretty frustrating!
Hi Andrew, thanks. It is probably good practice to stand on the little pistons, as I am sure the crankcase with integral cylinders is going to be a real learning experience for my little foundry :)
Past employ as GE apparatus service dividion and years later J Deere maint (shop worker), I am a new subscriber, WA state, US. I own a S Bend 9 of 1935 vintage. I am delighted to watch the progress to this curious design and sort of astonished as to the strict build to extracted plan. I would change up anything as found useful, you see. I repair various Honda superbike (1986) CB1100 as carburetor type. A 98 Jeep cherokee limited *(grey leather) 53k original miles and find myself cheering for your project.
As a young teen, I chanced to visit a site with steam and various machines of 1910 to 1939 vintage (Mogul, other fuel) thus appreciate vintage iron. Semi-retired, I
have access to a huge scrap yard with rather the time and inclination to fitter, weldment projects. Thanks for having me along. M.
For the hobby or professional I recommend Max Grant, Swan Valley (AU) Shop's channel. Also, IC Weld. Looking forward to progress.
Welcome. I do not know closely I can stick to the original, bit will try. I do like the idea of this being a replica and not a modern interpretation.
Yay! MOTO content!
I'll have to agree with your Sis, tools for tools are nice but in my shop everything has to justify it's existence. I'm a moto builder who skilled up in these areas to make parts.
Salivating for this series.
Hi James, thanks for your feedback.
That is a fantastic project!! Thank you for sharing!!!
Thanks fo watching.
I would suggest putting your sodium silicate cores still in the molds into an air-tight container, pull a vacuum, and then flood it with CO2 and let it stand over night, this will allow more of the sodium silicate to harden, making de-molding,and cleaning up the pattern cores easier.
Good idea. Thanks.
"The Knife and Fork Man" is definitely one of my favourite engineering books. Looking forward to following this series.
Thanks Alan.
Trust you to pick an easy one for the first project! Really looking forward to this one!! You could try a stronger mix of sand with more sodium silicate or can you get some holes right into the problem area to make sure the gas is getting to it? Redrup is one of my heroes!
Phil
Ps BSF into aluminium is not reliable. Royal Enfield among others used them and I have spent a fair time helicoiling them!!
Hi Phil, thanks. I did put those little gas channels in, but I think I need to move then down by about a half inch to get the gas where it needs to go.
I dont think there is much aluminium in this engine other than the heads and pistons. Need to take a close look to see if the timing gear covers / main bearing carriers are iron or alluminium.
Perserverance is all that counts. Looking forward to this series.
Yep, this project will need plenty of it. :)
Looking forward to this series!!!
Me too Michael :)
Wow, that is quite the project!
Thanks, I hope you enjoy it.
Awesome! This couldn’t come at a better time. I’m really wanna start a motorcycle engine build of my own!
Cool. What sort of engine will you do?
@ I wanna build an engine based on 1930s Husqvarna racing single 350cc with overhead valves. But base it around a Vw 1200 cylinder. Will probably try to use some existing parts, like cylinder, valve, cams
@@Studiolindholm Cool project. Please make videos of it.
2025 is going to a good year for watching your channel. Best of luck.
Thanks a lot.
You're a patient man, Mark.
Time will tell :)
Oh. Hell. Yeah. I'm totally down for this project.
Thanks.
looking forward to this build
Me too Paul :)
BSF is a torture choice, the best of luck :D. Have the plans and raw castings for a BSA B50 V twin that I've not started yet so kudos for doing the first step and getting a project going. Looking forward to watching you carry on this project.
Yeah, BSF is very fine. CEI is no better though :)
fantastic project looking forward to your first ride.
Thanks Robert, might take a while :)
This will be an interesting project and likely lead to a huge amount of content. I am already rubbing my hands at the thought of all of the side-projects that will come along with the series. I wish you well🤝
Thanks, back to normal. Yeah, I am learning a lot on this project already.
Maybe rotating the split line of the mould 90 degrees so the halves meet at the pin would give less surface for the little bits to hold on to?
Good point. I will see what it is like once I have finished grinding off the excess epoxy and maybe try that if nothing else works.
Dang, I was hoping for an axial compression jet engine build. Maybe next project. Looking forward to this build. And found you by the video for the doall bandsaw gear box fixes, I have the same saw. Cheers and really enjoy your videos.
Thanks for the kind feedback.
I helped make parts for aerospace companies such as Boeing, Korean Air, Liebherr, the Military etc. I’m into aerospace history just a tad. Can’t wait for this project of yours.
After this build you should consider making a copy of the Burt Munro 600cc modified Indian engine that he built for his Indian motorcycle, which still holds the record for the fastest Indian motorcycle. I lived in Utah when he set the record and I was about 10 miles away in a small border town called Wendover. I would have been thrilled to see the event in person, but I was just a very young kid in 1967. 😊
Yeah, that would be a cool project, but I would imagine that those engines are now rare as hens teeth, and it would be tough to fine either drawings or someone willing to let theirs be disassembled and measured.
Good day to you sir. I would really appreciate you replicating this motor. You have demonstrated your skill clearly in your previous videos. Your replication will help us understand how a man would approach this over 110 years ago. I couldn't think of a better man to undertake this project.
Your patience is remarkable for sure.
It is an interesting project. I learn something ever time I work on it.
Good book on pistons
It really is.
@@RotarySMP I got it and cupola furnace book off a bloke I sold some woodworking gear too. He had a few more out of print ones.
@@brendanshorter5550 Will you build a cuppola? Very simple but hard core given the huge rate of melting.
Really looking forward to the rest of this.
Me too :)
I basically came to the same decision you did last year after spending nearly 3 years just building my machines and tools to use on them, that it was about time I started using them to make things with. I had been watching a model steam engine build by Mark Presling and I decided I would tackle one of those, but he made one from a kit and I decided I wanted to build one from scratch with my own castings. However, my casting were only Aluminium and not Cast Iron like the kits, but I had no intention of ever running it on steam, so that wasn't an issue. After I started on that, one of my long-time viewers commented that in his opinion making model engines was the first sign of a dying channel, I just shrugged that off at first. During that build another long-time viewer suggested that I should tackle an IC engine next up, and I ran with that idea and have spent the last 2 months building a Webster Engine which I've nearly completed. However, I have been watching my view count steadily decline over both builds, and I'm now starting to wonder if the Dying Channel comment was kind of right, I sincerely hope that doesn't happen to you. On another note, to save a lot of time and hassle you should consider Lost Foam Casting, in which you can substitute the foam for 3D printed items, but they need to have hardly any infill so it will burn away easily, I have made a lot of lost foam castings and once you get the hang of it, it's a real time saver, and they come out pretty well. I just this afternoon ordered a 3D printer to give it a whirl as a week ago I bit the bullet and downloaded FreeCAD 1.0 and got stuck into learning that. I'm personally looking forward to watching this build. Sorry about the long comment.
Interesting feedback. Thanks. I was once going to build an Edwards Radial 5 model, but decided that I dont really want to do model engineering. Maybe I am kidding myself, that this is not a model build but an actual replica of a motorbike engine, so I would also have the possibility to continue on and build the frame and motorbike for it.
I sort of considered lost foam casting, but this engine is still small enough that I should be able to cast the crank case with cylinders conventionally. I will start making at least two sets of everything, and even more with pistons. I do like what Kelly Cofield is doing with lost foam though.
I want the multiples of parts to force me to set up CNC machine jobs properly, and not just dick around with one offs :)
@@RotarySMP My problem is I only have small machines, so building anything bigger than a model is not really feasible.
I did the opposite to you: I have the printer and need to learn CAD software
@ It is a very helpful skill.
@@alanclarke4646 It's the one thing that had been stopping me from buying one, plus a complete lack of somewhere to store it, so it looks like it will have to live on the dining table when I'm not using it, which won't please she who must be obeyed.
That looks like a great project 😀 looking forward to following this. Maybe a few clout nails in the weak part of the core would have helped?
Hi Julia, thanks. I am finding it challenging right of the start. It will certainly improve my casting skills.
Mark, it looks like you have no draft on your mould. There should be no vertical faces when you look down on the two mould halves. Did you add a shrinkage allowance to your mould? 😊
You got it. I only gave that pin boss and rib 1.5° which is not enough for such an enclosed feature. I'll redo it.
Yes on the shrink allowance. Nice thing about 3D printing. You just design the model nominal, and then scale the STL model 2% in the slicer.
I highly recommend to make 3D prints of the most complex parts to test that the interaction between components to make sure that the engine will turn over without interference. I’m sure that there are parts where the drawings are ambiguous, and holding a 3D model in your hands will help a lot.
Good point. Having only two views, there is plenty of information I don't have :)
Great project, thanks for doing this, I look forward to watching the whole series. Btw if you want to remove the facets on your 3d printed FreeCAD parts, open up the settings, go to the "Part/Part Design" section, then "Shape view" under that, and reduce the "Maximum deviation depending on the model bounding box" to maybe 0.02%.
Thanks for the tip, I did wonder why they were so faceted, but figured the sand would not care :)
I just went in and adjusted it. Thanks. Actually, I had gone looking for this setting earlier on, but had no idea what it was called and didn't fine it.
Hi Mark., Good luck with you journey. As it was already noted earlier these zones are not getting enough CO2 to harden. you can drill additional holes that will allow to harden this better OR you can use epoxy glue instead of calcium siliicate - epoxy will cure itself - no gas needed and it will cure evenly on all core.
The Silicate sand is great for quick iterations. I need to redo the core mould with more generous draft angles.
What an incredible project!
3 cylinder bikes are making a comeback, maybe you can help bring back radial bike engines too. 😁
Re the sand molding, looks like the strength of the sand is too low for the strength needed to hold those detail "ears".
I recommend really increasing the draft angle on your cad model for any small details like those ears. You looked to have almost no draft angle at all and with such a weak material (crumbly sand) you need a heap of draft angle like 12 or 15 degrees.
Another thing would be to increase the strength of the sand with more binding agent, making it less crumbly.
And as a final suggestion, I'm not sure why you are making that tiny sand core rod for the piston pin? I think the pour will work better without that complexity, less chance of air entrapment or premature metal cooling if you just make the piston with solid metal in that area. After all, drilling the piston pin bore is one of the easiest jobs, personally I would prefer to drill solid metal there and not try to drill in a hole full of sand particles and I think it will just mold easier.
But then maybe I'm just missing something and there is a good reason for the core?
Yeah, I had draft, but just went back in and checked my model and only have 1.5° there. Normally I use at least 3° on any internal surface, and as you point out, that little corners need more than that. I will revise the core model.
I guess I was just following the book, but also wondered whether that pin core gains me anything. Normally it would reduce the risk of a shrinkage error from large concentrations of metal, but this is a pretty tiny piston.
@RotarySMP ooh yeah 1.5 degrees is the kiss of death lol.
5 or 6 is a minimum for a weak material in a non compliant (ie; solid) mold, and then for small details you need to increase far above the 5 or 6.
Maybe you are overthinking the complexity needed in the piston core cavity? People make perfectly good pistons out of a slug of solid metal, just cutting out the conrod cavity with an endmill. I know you want to mold the piston, just saying maybe the core sand slug doesn't need to be that complex?
Anyway thanks for the videos, the process is as rewarding as the success, whichnis why we do this kind of stuff I guess 🙂
@@wizrom3046 Thanks, yeah, I am redoing this model and giving it more draft.
I still have to do the Same Thing , really start to make Something instead of Tools for the other Tools.
Looking Forward to this
Yeah, I have been putting this start off for too long.
Wax mold release on PLA 3d printed molds takes a few heavier layers. On composite parts I've done I've had to add a layer, wait a while for it to completely dry, and repeat at least 3 or 4 times.
Thanks for the tip.
I'm about to start down this path myself. I have a 1920's designed inline 4, water cooled engine that needs new oversize pistons, and the original pistons are cast iron. Not having a piston cam grinder to remake a similar size piston with proper skirt clearances, I am looking at either machining my replacement pistons from solid, or making the core molds as you're doing here.
I've had similar success with 3-d printed core molds, and a lot of the issue I had was layer line direction. I watched some videos by Easy Composites here one youtube where they were making 3d printed molds for carbon fiber parts and I'm going to try to implement some of their suggestions and methods.
I think I was just too tight on the draft in there. I only gave the pin boss and rib 1.5°. I'll remodel it with more.
What a project 🎉 very looking forward to it........ for your sand mixture: glucose is a honey like liquid and would glue the sand together like epoxi.......sucrose is just crumbly and just makes your sand sweet.
Hi Christan, it is not about the cold sand, it is about weakening the sand once it has been backed by the hot metal. You want that core to collapse, without the sugar, it will be a bit concrete like and hard to remove.
gloucose is like honey, it will spread more even and i think it will caramelize and burn to carbon when you pour and remove all adhesion so it will cruble quite easy afterwards........ just my expirience from cooking 😅
@@SuperAnodyne Sucrose has the same effect.
Looking forward to seeing you progress through this project, very exciting! On the piston casting, could you rotate the orientation 90 degrees so that the wrist pin features are split half to each side for better sand plug removal?
I think the issue is lack of draft. I just checked and only used a 1.5° draft angle. No where near enough there.
I think you should prime and polish your mold and put a pin hole in each of those corners that break out. Additionally if your CO² dispensing shroud was slightly concave and covered the entire mouth of the mold it would help force the gas down into the corners.
I havent done any casting since highschool, but i dont think there's any way that raw plastic from the printer is going to release cleanly.
Hi Jim, I am really surprised it releases sand as well as it does. With my old printed I had to fill and prime to remove the striations, but on the Bambu I am printing at 0.08mm layers and it has very little striation. The big issue turned out to be draft. I rechecked the model, and had only given the pin boss and rib 1.5°. I redesigned it and reprinted it, adding more gas holes, and at the same time printed a full size CO2 shroud (I was meaning to already). On a work trip today, but looking forward to trying the new mould. It would be nice if I can get away without painting and polishing it. Obviously if it still hangs up, I will have to.
Wow this looks like fun.
I've found that UV cure epoxy is an excellent alternative as you can build up a structure gradually and exactly where it's needed layer by layer. A kind of manual resin printing if you will.
Also, as it is the core, would making it in sections with a final glue up work? ie expand the whole gudgeon pin core to include web and fillets and then insert it into one larger hole in the piston wall.
Could work, but I suspect I am pretty close now, and only need one more session with the die grinder and this core will be serviceable.
Did you forget to add draft angles around the area of the 3d printed mold for the sand core since it kept refusing to release? didn't seem like you had any from the video
Nope, they have draft. At least had it before I started smothering it all in epoxy.
@@RotarySMP was going to be my suggestion also...maybe the bosses need more draft? The sand is quite trapped in that area
@@ollied1981 You are both right. I rechecked the model, and only gave it 1.5° there. That was no where near enough for such an enclosed feature.
good video RotaySMP
Thanks for watching.
Can’t wait for this!!!😀😀👍
It will be a slow journal. It only makes 3HP :)
It’s not the destination, it’s the journey😀👍👍
@@motorrad1527 It is. I am learning every time I spend a few minutes on this project.
A proper mould release is obviously the thing to use, but I have had good results with baking spray “spray and cook” on epoxy
Thanks for the tip. I think my main problem is insufficient draft angle.
Happy new year. This is going to be a fascinating project very much looking forwards to it with all of its ups and, inevitable downs.
Thanks.
you got me sub within the first 15sec
Welcome and thanks for joining.
I am looking forward to this build. Any value add to placing the blanks in a vacuum chamber then applying the CO2?
Thanks Brian. The sand is so porous, it should be easy enough to get all areas in that little core gassed. I probably just didn't hold the pipe there long enough.
Glad that you're apparently over whatever virus that you had. Judging from the beanie/hoodie/sweater (jumper) combo it looks like your basement is not any warmer than the 13C that you cited previously.
Hi Brian, thanks, I am fine, but yeah, that is the temperature in the basement till spring.
Ohhh wow!!! This project will be one to follow!
BTW, have you looked into using epoxy for the core moulds? Mark Presling has used it successfully several times.
Epoxy core sand could make it worse, as it looks like a lack of draft. An epoxy sand core would not break, and could destroy the mould. I need to redo it with more draft.
A Megola rotary would be interesting...if not more challenging to build.
I knew the British had several systems for threads which were largely incompatible but I didn't know they had one just for bicycles and motorcycles.
I remember rading about them. Looks like a suicide device to ride.
Yeah, CEI (Cycle engineering institute) threads became BSC. I think it is something like all threads haveing 26TPI regardless of size, or something like that.
@RotarySMP Sounds like they would be very strong fasteners.
@@douro20 Until you cross thread them all as they are so fine :)
"Knife And Fork Man" bought. I must be the first person in the US today to buy a copy :) Being this early helps!
Nice one Vince. I have asked Bill if he has a link to a current publisher, as my copy was published by Diesel Publishing in Badminton, but I cant find anything about that. Once he gets back to me I will add another link.
I just ordered one as well.
@@alanclarke4646 It is a good read, and Bill is a very nice bloke.
Three random thoughts:
1. How much draft do you have on the wrist pin bosses? Maybe not enough, so that slight side to side twisting as you remove the core cracks it?
2. Maybe the sand isn't completely curing down in that part of the mold? Maybe a couple of CO2 injection ports right over the bosses, rather than just one at the bottom (or top, or whatever) of the mold?
3. If all else fails, print the mold out of something somewhat flexible, so you can peel it off the core?
1. Insufficiant. I rechecked the model, and only gave it 1.5° there. Need to redo it.
2. I have three CO2 injection ports , and will move them even closer where they need to be.
3. Good idea, a little less rigidity would also help.
I have three gas canals to that area, but can probably optimise their location.
Once I get this core mould dialed in, it is pretty quick to cast a whole batch of pistons. It is no trivial to machine those voids. They are deep, narrow and have undercuts.
This looks like an excellent project that I am going to enjoy watching. Looking at the problems you are having with your cores, I would hazard a guess that the problem is a combination of not enough gas exposure in those areas of the core box and also that the sand might not be packed firmly enough...
Yeah, I can see that some of them I packed better than others. The silicate sand doesn't pack down the same way green sand does, but still needs more packing than I gave it.
I wonder if you are getting gas to the spots that are crumbling rather than it being a shape problem.
I am pretty sure it is a shape problem Robert as I rechecked the model and only have 1.5° of draft angle there, although on some of the attempts I might have undergassed them,
I expect your CNC milling machine can be used to clean up a looser casting but they did manage a solution in 1920. It might be worth sending a DM to olfoundryman and seeing if he will answer.
@@robertfontaine3650 Oh Martin and I have been in close mail contact for the last few weeks, and once he sees this video (middle of the night in Oz), I am sure he will get back to me.
I screwed up the draft. I checked the model and only had 1.5° around the pin boss and rib. I need to redo that.
Very nice project! Looking forward to see more 🙂 I don't have any knoweledge about sandcasting, but I would try to split the core form into 4 instead of 2, to reduce the stress when you pull it apart.
I hadn't given it enough draft angle. A new mould was printed yesterday. If I do need to split the core mould more, I will slice off the bottom and make it removable.
First ? Good Morning, Mark. This sounds like an awesome project! I especially love your casting videos.
Hi Vince,
Thanks, then this series will be good, as there is going to be plenty of challenging casting.
Would using a three part mould for the cores, with a separate bottom piece, be any better? My thinking is that the less surface your trying to separate at once might help prevent breakages. Just a thought.
That could also help, but the insufficient draft angle (I only gave it 1.5°) seems to be the main issue.
I couldn’t click like fast enough. I’m excited for this.
I like the getting started approach. I wonder if painting the prints with a high sheen paint will help with release.
What draft angle are you using? For one of my latest patterns everything has a 2* taper. In fusion it’s handy that you can add draft to any extrusion.
I rechecked my model, and had only used 1.5° draft angle, which is too little for such an enclosed internal feature. I will redo it with much more. FreeCad also has that option.
Hey Mark - I am looking forward to this series :)
Hi Nikolai, thanks for dropping by. :)
Decades ago when I ran a printing press, I would polish the back cylinders with Johnson’s paste wax and then spray them with silicone spray. Perhaps that might work for you as well?
I was also thinking I should get some silicone spray. Thanks.
Omg, this is so me. I work in IT but hobby have constantly improving my not small workshop anymore. I had always squint when getting this question: "what have you actually build" 😂. Nothing that I can think of is the answear. Im also thinking about building small engineering myself from some time but didnt get to the actual start part yet 😅
Go for it :) At least we will than have an answer... a never ending engine build :)
Sounds like a cool project.
The dual camera setup in the beginning was pretty distracting
I was trying something new. Normally I dont do talk to camera, but had nothing else visual to use. Hopefully the further videos will have more to show.
Yep, no worries, just feedback. Hopefully constructive?
@ It was thanks. I noticed when I edited it that I was watching the screen just off the camera axis, rather than looking at the lens, and also that I screwed up some takes and looked at the camera which was off (I got out of sync of starting and stopping the two camera and audio recorder, and ended up with some clips the camera recorded between dialog. Duh! :)
With a face made for radio, I prefer the videos to be about the project rather than showing me.
I liked the "Something new"
To be clear, I wasn't concerned about the talking to camera, that was good, just the transitions between the two views.
And in the meanwhile, the only Belgian participant Denis Van Weynbergh is last. But he didn't gave up. Go for it, Denis.
Hi Michel. It is pretty wild that some of the field was up to half the globe behind the leaders.
Free Cad fillets, and the pain. I'm always incrementing them in 0.2mm until it breaks, then back off a little. Failing that, create the mirror profile, and do a boolean cut
I do the same. I am getting better at FreeCad :)
Could it be not enough C02 getting down into the bottom... maybe some extra vent holes down in the fillets? Happy Sunday!
Thanks Mike. I have 3 holes there, but will relocate them.
Have you considered something like an ejection pin/ejection surface for the difficult to cast areas in your sand core mold? 🤔
Not yet. I am pretty sure a little more grinding and polishing in that area and it will start working reliably. Add ejection help may be an option if needed.
Using real mold release is lava.
Re: the filets - it's the reason why Parasolid and ACIS are the most popular modeling kernels, since they can accurately compute filets intersecting with each other. If there's no particular reason for filets, you could try chamfers, since they are geometrically simpler to calculate.
Hi Alexander. FreeCad seems to hit a limit when a fillet would touch another feature beyond a certain angle. It is probably helpful for me to do this with epoxy and grinding, as I dont really want to keep printing waste plastic core moulds which dont work.
@RotarySMP Yeah, it's a foundational problem with the modeling kernel. One of my colleagues is a beta tester for Plasticity, he also makes tutorials here for that software, and mid-development, the developer switched from some kernel to Parasolid, since it caused to many limitations in what the software could calculate. Including filets and chamfers. It seems like a solved problem, but in reality relies on kernels made 50 years ago, that to this day are getting commercially licensed, without much chance for open-source software being able to reproduce it.
@@graealex Freecad is certainly a victim of many of the weaknesses of open source software, but for the most part I can make it work for me, and it is improving.
@@RotarySMP It's going to be hard to overcome this fundamental flaw. The big CAD players' shopping sprees made sure of that. And fillets aren't just a cosmetic thing, they are integral since real-life tools seldomly create sharp corners.
@@graealex True, but this old engine is also not rocket science. If I can get close with Freecad fillet and a 3D print, it is not a deal breaker for me to go old school and add fillets of epoxy here and there.
I got the feeling, it breaks where you have low flow rates of CO2. Maybe try vent holes right in the middle of the stuck part or leave the whole thing in a nice warm CO2 bath over night. Anyway, wish you success!
Thanks for that tip. A bucket of CO2.
this might be a good part to make with sacrificial plaster molds, the kind where you 3D print the core, embed it in a plaster mold and melt it out to make the negative space
it does mean making 3 molds, one for each piston, but it might be easier
Oh, you mean investment casting. It is not really ideal for this kind of task, as you have to burn out the investment (takes about 12H). It is great for thinks which require accurate geometry but not engineering fits. In this case, yoou have to machine the OD, the ring grooves and pin bore so it is a waste to investment cast it. Also I dont need three, I need enough piston blanks to end up with three good pistons after inclusions, machining errors etc. :) Once I get this core mold dialed in, it will be much faster to sand cast a batch of pistons.
@RotarySMP yeah is probably much better for one-offs or complex geometry. with proper draft angles the pistons should be very doable with sand, although the housings themselves might be a different story when you get that far
@@terjestrm1796 I started modelling the crankcase with it's integrated cylinders. That is going to be a real challenge. I have also modelled the cylinder heads, and found them more difficult than expected. By the time I am ready to try casting the crankcase, I whill have a bit more experience in both casting and 3D printing core moulds and patterns.
question: was doing a lost PLA molt not attainable? (printing cylinder head positive with PLA, pack it into the sand, and bake out the PLA at 240C in the oven?
Investment casting is done with a plaster like material, and it takes 12hr to burn out each form. It could be used, but is a waste as investment casting gives very good accuracy, but not enough so you still need to turn the OD, ring grooves and pin bore.
Once I have this core mould dialed in, it will be very fast to produce a batch of pistons. I need more than 3 as I'll have some with inclusions, and screw up the machining on some.
Book bought. *Really* looking forward to this series! Especially seeing your amazing machines in action. I think (being the expert I am (not)) increasing the draft angle of the gudgeon pin wall and its web (to the underside of the piston) would help more than increasing the fillet radius?? Also easier to do in FreeCAD ;-)
Hi Philip., yep, I rechecked the model and had only given them 1.5° I'll redo that core mould.
Hi Mark interesting new project. a suggestion from someone that knowns next to nothing about casting and cores.
what if you prepacked each half of the mold first to ensure the sand around the bottom is truely compacted.
Hi Steve, this silicate sand doesn't compact down like green sand does. It sort of just moves around. I do need to compact it a little better, but am pretty sure that is not the main issue. I need to smooth out that epoxy fillet I added.
This is going to be good :D
I hope so as well.
Just a thought, why not model tricky bits as a separate printed insert so fillets can be modelled in better draw situation??
Good idea. Hopefully I wont need it, as I put in insufficient draft. I'll redo it.
have you tried playing with drawers for your core mold? they are super expensive when injecting plastic, but you are demolding your cores by hand, you can pull on a piece of plastic for the same price. I would probably try to make the boss around the drill bit a drawer, pulling it towards outside should shear the bond between the mold and the core, and when opening the main mold the sand will be far less "trapped".
Not sure what you mean my drawers? Do you mean guide rails so that it pulls more parallel?
@@RotarySMP sorry, I used the wrong word, it looks like it's called "SIDE CORING OR SIDE DRAW PINS" it's like your drill bit, it's a separate part of the mold that you have to pull before opening the mold because it's not demouldable by straight pull. Basically I propose to 3D print your drill bit+its plastic shoulder in one piece (split in 2 at the parting line), and pull that first, then open the mold.
My reasoning is that currently when you open the mold you have to shear 2 molded faces at once, by having a side core, you shear one face when pulling the side core, then the other face when opening the mold.
@@RotarySMP I just published a video for you
@@nraynaud Thanks. I have an email address on the channel main page. Would you be able to send me a sketch, as I am struggling to follow the description.
@@nraynaud Thanks a lot for sending that.
Perhaps some graphite to help with the corners. It was a common practice to use a water based spray with graphite in the silicate core dies. However they were usally not as detailed.
Thanks for the tip. I tried to find a spray bottle to do that with a while again, and found the nozzle would clog before it was useful. I need to find a solution for this before I get to the main crankcase iron casting, as that will be a bear and a graphite wash would probably really help.
@@RotarySMP We used to make simple air driven syphon guns. The crankcase may benefit with a core paint. During low pressure aluminum casting the naked cores would get worn and distribute sand in the casting. The core paint works well in high flow areas.
@@dazaspc Thanks, I need to get that sorted before I get to the crankcase.
Brilliant! Looking forward to the rest of the build, I'd love to do something similar in the future
Looking at those ears flaking off, do you think a few more gas holes might ensure they cure? Or would it make sense to cure halfway while the core is shallow?
More draft and better located gas holes are my next iteration.
@RotarySMP fingers crossed it goes well!
@@mrsockyman The next mould is on the printer now.
I'm excited to see this project gain momentum! The piston ring retaining the gudgeon pin doesn't make a lot of sense, you have all the extra friction of the piston ring but it doesn't provide any sealing benefits as there is a direct path around it through the gudgeon pin hole. Is the lower ring supposed to be an oil control ring (doesn't appear to be)?
From the drawing all three rings are the same. I wonde when oil control rings were invented and who patented them?
Could you get the mold in a vacuum chamber? Removing the air then adding CO2 would eliminate any soft spots. A pressure cooker could be modified as the vacuum chamber.
Probably could, but this sand is very porous, so it should gas fine if I am consistent with the duration and ports I gas it from.
In plastic injection molding, the mold release we use contains PTFE. Not sure where you could obtain it where you are. Why are you building sand cores rather than using lost-PLA casting?
Thanks. Once I get the core mould dialed in, sand casting is way faster. Investment casting requires each mould to be burnt out for 12H, and the investment is expensive.
Oh this is gonna be so interesting, thank you Mark. The sand cores, when it has been used for the casting, can it be broken up and used again or does the Sodium Silicate (?) totally mess with the sand?
Typically you try to separate the core sand from the green sand during shake out. A bit of core sand will get into the green sand though, but I dont think it makes much problem. It will just get mulled into the green sand. I'll just dispose of the separated core sand. A 20kg bag of sand only cost a couple of € from te hardware store.