Thank you so very much for this. My perfectionism did not allow me to settle with any of the countless videos that show how to model "gear-like objects" that dominate the search results when trying to learn. This tutorial was perfect. A proper approximation, brilliantly explained and demonstrated. I have learned so much from this and will be referencing this series quite a lot, I'm sure. Thanks again!
Thank you! This episode especially was a bit special to make in a way, because the first videos in this series were mostly made because existing tutorials were usually incomplete. For this one though, most other tutorials are just plainly wrong.
I realise you published this ages ago, but still wanted you to know. between this and your worm gear video (the only two I have watched so far), I have learnt far more about fusion 360 than I have in the past 5 years of hobby use. That's despite watching a number of other tutorials and countless hours or googling.
I have to add my voice to the choir: Simply outstanding! Honestly one of the most educational, interesting, and USEFUL series of videos I've seen on UA-cam. THANK YOU for the thoroughness and fantastic presentation. I'm just sad it's taken me so long to discover these!
This video saved my butt. I was trying for days to model a bevel gear set for a project, and the mesh never worked out. I didn't understand about the back cone and virtual gear. Your explanation was perfect. I am using SolidWorks, its wild how different F360 is.
Yeah that's definitely possible, but it can be tricky to put the bearing there and still make the whole thing assemblable. For high torque or high power applications it's worth investigating.
You can first push the button for the parallel constraint. It may not be on your ribbon bar, but if you click on the word "constraints" you get a dropdown from which you can select it. Afterwards if you select both lines and use the dimension tool, it will set the distance between the two lines. Usually it will set the angle, but it will not if you've already defined the lines to be parallel. I almost always use the shortcut D to use the dimension tool.
Very clear and complete instructions, as always! But far more complicated as plain spur gears, and placement of the gears on the axes is quite critical as you explained. I want to 3D print pretty small gears for use on perpendicular axes; about 10 mm diameter for the pinion to mount on the axis of a small Mabuchi type motor (as the ones in servo's), driving a large gear of about 80 mm diameter. After seeing your excellent tutorial, I'm now considering to make thinks easy for myself and use a regular spur gear for the pinion and a crown gear for the large gear. But crown gears don't seem to be very common, except for in toys, tower clocks and wind mills .... How do they compare to bevel gears?
I'm still a bit confused by a crown gear to spur gear mesh. The way I understand it now is that it can only work with copious backlash and some friction. I probably just don't understand it correctly. The wind mills of old didn't have the technology for proper gears of any kind, let alone bevel gears. Modern turbines don't use crown gears as far as I know. I think your problem can be done if the torques aren't too large, by SLA 3D printing. The pinion will be very hard to do with FDM I would imagine.
@@antalz Thanks for your prompt reply. I'll do some experimenting. I just found 2 articles on this subject, you may find these interesting too: www.geartechnology.com/issues/0901x/Weck.pdf www.powertransmission.com/issues/1013/crown_gears.pdf
I'm afraid not, I think I came up with a modelling method at some point, but I was never confident it was correct. UA-cam is full of incorect gear modelling videos as is. I also forgot what that method was exactly.
@@antalz I really appreciate all the work you've done on these videos, what you have to say about there being too many incorrect videos is unfortunately true. Thank you so much for all your help and guidance!
What do you think would be the biggest load of a linear actuator propelled with 3D printed gears and what would be the best gear layout for durability. Thanks
My force-extension machine can produce about 300kg of force, but most of the heavy lifting comes from the M12 threaded rods and nuts. Still, there's no real upper limit if you make the gear large enough.
@@antalz Thanks for that Antalz. Would a proper ball screw be better than buying an M12 rod from a hardware store. For your project, would a M14 bolt work better ? If you were to use your machine to move a 50 KG load in the vertical plane, with a stroke of , say 15 cm, and a period of , say 5 seconds, ( 3 cm a second ) , what would be the best combination of threaded rod, plus gears. 3 cms a second may be a bit quick for a normal M12 threaded rod. Sorry, I don't wish to be setting you a task here. May I re-phrase the question to a simple yes or no. WOULD IT BE POSSIBLE USING 3 D PRINTED GEARS AND AN 895 MOTOR TO PRODUCE A LINEAR ACTUATOR THAT CAN MOVE A 50 KG LOAD OVER A 15 CM STROKE, IN A PERIOD OF 5 SECOND ? A simple yes or no answer would help, thanks. very much.
@@jackspratt4343 Yeah that should be possible. What you're asking is 500N at 0.03 m/s, which makes for an output power of 15W which most 895 motors can easily put out. I'm not sure what setup you should use though. Probably a very low or no gear reduction on the motor. If you take an M8 rod, with a 1.25mm pitch, you need to rotate that 24 times per second which is a bit much. You can use bigger rods, because they have coarser pitch. You can also use TR8x8 trapezoidal screws, I'm quite sure they can take that load. For a TR8x8 you'd only need about 4 rotations per second, so that would require some gearing. I recommend buying a motor with integrated planetary gearbox in that case. It's probably better to couple the motor to the rod directly using a flex coupler. Do use the ones with orange plastic spiders, not a fully aluminium one. No need for 3D printed gears there. It's also very hard to attach a gear securely to TR8x8 rod, or M8, it's much easier for bigger rod like M12 and larger. Basically, find a threaded rod that has coarse enough pitch that you can obtain the desired speed. For my own device, an M14 rod has coarser pitch, which means my machine would move faster, but exert less force. Even low-grade M12 rods can carry several tons of load, so that's not a concern.
@@antalz That's great advice, Antalz. Thank you, May I ask whether the power calculation, of 15 Watts for 500 N @ 0.03 m/s would "extrapolate" by a crude rule of thumb, to 30 Watts @ 0.06 m/s and 45 Watts, in an ideal world @ 0.09 m/s.? Thanks again.
Thank you so much for making these videos! This is only of the only places on the internet to find all the correct info in a very well explained and complete manner. You deserve so many more subscribers. I'll do my best to share your videos.
I've been waiting for this one and needless to say it's a masterpiece! Your videos always feel like they are only 15 minutes long that's how gripping they are. btw your insta link in the description is broken :)
Fantastic video, I am learning a lot from this series. I would love to see your approach to making a helical bevel gear in Fusion. I'm not quite sure how to get the appropriate twist in the lofted teeth, and how to get them to mesh properly. That would be an interesting video.
Thank you! I thought of doing it using a loft with guiderails, but I haven't actually done it. I think you will also need to emboss the curve onto the pitch cone, but again still need to look into that.
@@antalz I tried with guiderails with both Sweep and Loft features, but I couldn't obtain a satisfying enough result. So I ended slicing a straight tooth, rotating each slice as needed, then executing the Loft feature with all slices as profiles.
Its things like this that we take for granted till we have to do it and then you realize the immense amount of math and things you have to wrap your head round that goes into projects/products we see every day, it just blows me away. I never knew there was so much to learn about gears, it always seemed straight forward to me till now. Thank you for this video!
Hi! This video is crazy, it literally teaches very good, but i still have a problem unfortunately… I don’t really get how to calculate the dimensions that i must input to work well but to be much smaller than what you did. I know u published a very long time ago but i hope you will have time to write a quick explanation! Thank you:)
I don't like leaving warnings in my timeline. And whatever I design is nowhere near as meticulous as this. Leaving a warning in the timeline on your design would be like inventing a perpetual motion machine and not starting it. Still trying to wrap my head around this one as I am going to need it. Will need to do some dry runs before incorporating into the main design. Great job explaining it. I don't think I can find anything remotely close elsewhere.
Thank you very much for this video and in general for all the others. I knew this construction method mechanically but now I know the reason for it. In addition, the way to apply it in F360 is brilliant.
This is excellent - I used your method with Solidworks (and manually modelled involutes for the fun of it) and it works very well. Thank you :) I must also say, that I am envious of Fusion 360 being able to animate parts and examine motion without having to swith to an assembly file. That is very cool...
Okay, I spent a couple hours trying to take a similar approach in Missler Topsolid, where I don't have a gear generator to lean on. It does have an involute curve generator though. I got to the point where equal teeth on the 2 gears mesh perfectly, but once I shrink the pinion, the teeth no longer mesh. If you have a few minutes to look at this video, that would be a huge help! ua-cam.com/video/6ZFHzi5EphM/v-deo.html
hello,this kind of drawing won't make the bevel gears conjugate, as you say approximation. and we can make the real conjugate bevel gears,but not your method.
Do you know a modelling method for true conjugates? I don't think it matters much because the approximation has smaller errors than almost anyone can reproduce on their printers. But it'd still be interesting. As far as I know even commercial producers use one of a number of approximations to ensure they can mill the gears.
At 4:40 - You mention the line is "1.25 times the mod" , in this case you entered "2" for the mod. what is the mod in this case? It seems like it is the tooth depth, and if so, why not call it that instead of a "mod" ?
The module is a measure for how large the gear's teeth are, and is related to the pitch diameter. In this case the bevel gears we're building are module 2 mm. Now when measured from the reference cone (compare reference circle or pitch circle for normal gears) I want the teeth to have an addendum of 1*module, and a dedendum of 1.25* module. The gear generator will take care of the addenda, but for the dedenda I must specify here that the cone that makes up the body of the gear is 1.25*module smaller than the reference cone. For bevel gears you essentially do all the math for the outside edges of the cones, and then from there loft everything inwards to the apex. This also means the teeth are larger on the outside than on the inside, so when we say "module" for these teeth we mean the module on the outsidemost end of the tooth.
Yeah that should be possible too, it might even be better because you can get that fractional virtual number of teeth. Drawing an involute manually is a chore though.
I am not the person who is making comments etc. But you really deserve Bold and Huge Thanks, my friend. This video and the thing you shared is tremendous. It has never been easy this much. Thanks for being clear, and helpful, and for your greatness. I am a huge fan of you after all!
This is great but I tried it for a 30mm diameter, so radius would be 15. Approximation is around 21.25 so I take it 22, but the gear tooth stays to outside for module 2 gears. 21 also stays to outside...
Do you mean you're getting interference? This might be because if you're doing 15 teeth undercut is required, which the Fusion360 gear generator doesn't do correctly. Even then undercut is better avoided if possible, a larger pressure angle may help.
@@antalz Hi there, thanks for the answer, I will try different pressure angles. No, I meant when you align the single tooth with the plane you created along the intersaction axis, your gear tooth is almost halfway embedded inside the cylindirical body. In my case its not embedded, kind of touching in bottom middle, but edges are lifted. Maybe it is normal and I need to sketch and extend so I can create the solid feature, but it feels like it is wrong and I might have problems later on.
@@ROBOROBOROBOROBO If I remember correctly, the tooth getting properly embedded in its gear occurs one or two steps later when the base of the tooth is push-pulled a bit to embed it properly.
At the moment no, I don't know for sure how to do it. I had a method in mind, but I'm still not sure whether it's legitimate. UA-camr "steel stone" has a video: ua-cam.com/video/L_hcaxxKt0g/v-deo.html
I am getting small amounts of overlap on the meshing of the teeth when I make gears with a higher module even with backlash, I do everything else the same except in the case of the gears I am trying to model they are a ring and pinion with 43/13 ratio and a module of 3.5 and I am using a backlash of 0.1, should I just move the gears further apart to force separation like in your pt. 2 of the gears series?
If the overlap is small I suspect it's because of the 13 tooth gear. That is below the undercut-free limit of I believe 17 teeth for 20 degree pressure angle gears. Either increase that tooth count, or a pressure angle of 25 degrees should work. Fusion's generator can not model undercuts. Also check you have a correct combination of gear ratio, shaft angle and cone angles.
@@antalz oooh thank you very much! I suspected that I should increase the pressure angle as it's a high torque application, this model is for a transfer case. The overlap is very small but visible by eye so it would bind in a 3d print for sure, maybe I could manually add a small undercut for low pressure angle gears with a small number of teeth? I believe my combination of ratio, shaft angle and cone angles is all good, my sketch looks to be correct and I used simple formulas like you did but with my 3.5 modulus in place, the joints and motion links also work correctly minus the small overlap which is consistent on every tooth so positioning seems correct. I'll try the higher pressure angle and if that doesn't work I'll manually undercut the teeth, thank you for your help!
@@antalz After a bit of investigation I found that even with a pressure angle of 25 degrees I had a slight overlap of the teeth, when I had the backlash set to 0.1mm I found that moving the pinion 0.4mm further away from the ring gear allowed the teeth to mesh well but technically now the pinion would be off center of the correct position of the shaft, would this have a significant impact on the performance of the gears?
Hi, great video I'm interested, what would happen if instead of using the gear generator, you make the tooth profile for the gear with 25.73 teeth or any non integer number. Would it be reduce backlash between teeth, or it would cause them to interfere? Thanks for your time
I think you could in theory make a slightly better fit with that, but the profile difference between 25 and 26 teeth is very small. Very few people are making bevel gears at such precision that it matters.
Tx for sharing that completed gears lesson. Btw, you told to manually adjusting the layers height within the cura's slicer, with bigger at base, thinner at top ... but how do we do that ?
I still get confused when to be working in the original part or the separate components. When the gear tooth is lofted and trimmed, you have the bullgear active, but when the teeth are patterned, you go back to the original part to pattern them. Why aren't they patterned with the bullgear active?
I went back to the root component there so I could edit the big initial sketch. The thing is, if you do an operation for "new body" or "new component", those get put into the component that's active at that time. So for the initial loft I'm doing a new body, so the bull gear must be active so that body ends up in the bull gear component. For cutting and patterning it doesn't matter, those features end up in whatever component you were cutting, or in the component that held the thing you were patterning. In fact a cut can affect multiple components. If you look on the timeline you see that the pattern and cut both ended up with the color of the bull gear component.
hey Antalz I love your tutorials they are very informative! I was wondering if it wouldn't be too much trouble could you further expand on this tutorial by adding how you could create a helical bevel gear?
Thanks mate! At the moment I don't quite have time to make videos unfortunately, but I'm hopeful I can get back into it. The bigger problem is that I don't know how to create a helical bevel gear. I have some ideas but I don't know if they create correct gears.
@@antalz I'm still playing around with it but that's my problem as well. I'm thinking that i need to use the contact lines length to drive the twist angle. I then made a offset plane at 50% of that and again at 50% of that as well and then divided the twist angle accordingly and rotated the original loft sections by those angles and created a new loft
Are there any reasons to not have the contact lines/inner diameters of the two cones parallel to each other? Wouldn’t that offer more contact area? I may be wrong, just wondering.
No that just doesn't work geometrically. As you get closer to the apex, the gear ratio and number of teeth have to stay the same. To make that happen, the teeth have to become smaller as you get closer to the apex. For that to happen the inner diameters must converge onto the apex point, meaning they can not be parallel. Maybe you mean making the teeth longer, but that doesn't actually increase contact area; it just makes the teeth more likely to break off.
This is a good approximation. The spur gear used to give the tooth profile is normal (at right angles) to the pitch cones line of contact. I've seen where others set the spur gear rotation axis equal to the bevel gear rotation axis. This causes loss of tooth depth but not tooth profile width and therefore a very distorted bevel gear form.
Yeah that's a big reason why I made this video in particular. There are a lot of videos out there on gears that are wrong, but specifically on bevel gears there is a lot of misinformation.
@@antalz Thanks. I've been producing gears for decades on my CNC lathe and milling machines. Only ever spur gears. Now with 3D printing, I'm getting into other applications and one job requires bevel gears. So where do you go to learn such design? LOL. I love UA-cam but as you say, there's plenty of bullshit out there. I'm an Aussie. We don't say "misinformation" That's way too polite. LOL.
@@antalz understandable; amazing videos in the whole series anyways. I have little note cards sitting with the formulas you explained. Put a lot of perspective on the caveats and variables of desighning gears.
Not sure what you mean. The blue gear is attached directly to D-shaft of the motor. Between the pink gear and the bearing I put a small 3D printed spacer.
@@antalz i mean to say, the motor you had used has shaft diameter 5mm but in the gear hole is of 8mm...that black part top of the motor is extra part? To increase the torque or it is the part of motor? Or can you please provide another link where i can find the same motor ...i don't know but that link has some trouble 🥲
@@kamrealamshaikh4375 Yeah the 775 motor itself has a 5mm output shaft, but the motor I bought comes with a planetary gearbox included and attached. I think I used the 24V i13 version here: nl.aliexpress.com/item/32822227433.html I used the same motors for my tank project, which I still need to finish.
@@antalz yups, that's what i want to know....Thanks, great work by the way. I am using your design in my project that's why I need to know about that gear system 😉😉😉
Excellent tutorial. One quibble, though. At 37:30, if the tooth profile is the (almost universally-used) involute form then, even though the axes are displaced, the teeth still roll against each other without sliding. It's just that the pressure angle is increased.
I didn't know that, thanks. Do you have any source where I can read more on this? Does this relate to profile shifting, because that's another thing I should learn more about.
@@antalz The fact that I mentioned is at 4mechtech.blogspot.com/2015/06/comparison-between-involute-and-cycloidal-gears.html. If the gear teeth are too narrow at the base then increasing the pressure angle makes them wider at the base, at the expense of an increase in separating force along the line joining the gear centres. If you Google "involute profile shifting" and choose the Marples Gears site there is more info on profile shifting. However, I still don't understand how a profile-shifted gear meshes correctly (i.e. rolls, not slides) with a non-profile-shifted gear.
@@chrisarmstrong8198 I'm not sure that site says the contact remains rolling. It does say the velocity ratio remains constant, and I think that's correct. I also think you're correct that the pressure angle increases because a further-out piece of the involute contacts. I'm just not entirely convinced on rolling contact, though to be fair it took me a while to see it's perfectly rolling for a perfect involute set. I'll study this some more, thanks again.
@@antalz I offer this hand-waving justification :) Consider two rotating cylinders, with parallel axes, joined by a piece of taut string marked with equally-spaced dots and arranged so that, as the string winds onto one cylinder, it unwinds from the other. The dots trace out the involute flanks of (one side of) imaginary gear teeth on BOTH cylinders, with the dots themselves being the contact points. (To trace out the other half of the flanks, the string needs to be wound in the opposite direction on each cylinder.) The relative velocity of the contact points on two mating imaginary gear teeth must be zero since they are produced by the same dot. Changing the separation of the cylinder axes does not change this fact. Zero relative velocity at the contact point implies rolling contact.
wonderful series (I just stumbled across it), I would love to see an entry on cyclodial gears as well (given that in the planetary gear episode you talked about doing compound gears for very high gear ratios), both single stage and double stage (which eliminate the pins and bearings for the output shaft) do you have a patreon account or similar to allow us to support your efforts?
Cycloidal drives is actually a good idea, I have seen a number of videos of people who made them on UA-cam, the latest being James Bruton. However I do believe most of them don't talk in-depth about how they designed it, or what things to consider. I haven't yet opened up a Patreon or anything like it, but it's something I'm starting to consider. Not entirely sure what to do with the levels and perks though. AvE seems to do quite well with his single-tier beer fund, but most creators have some progression of tiers and perks.
@@antalz re:patreon, I sponsor a fair number of people and mostly ignore the perks, start off with a beer fund approach and if it turns into real money, you can add perks that make sense based on requests (you have a business, so any perks you offer need to make sense from a business point of view) Applied Science does per-video sponsorship so he doesn't feel pressured to get a video out on a particular schedule, which is good for in-depth technical videos. re: cyclodial gears, yes I watch a lot of videos from people who are making them, but other than the gear ratio, very few people talk about the considerations in designing them (especially for 3d printing) what is the real effect of different eccentric values (is it like pressure angle)? I see some people with bearings on the outer pins, others without and talk about it being a rolling contact, what is real? almost all cyclodial gear setups are N:N-1 tooth ratios, what if you did something different? dynamic balance of cyclodial gears is an interesting topic, but I think it only matters if you are talking about higher speeds where the disks are heavier. If you talk about slow things (camera mounts, 4th axis for CNC, etc) this doesn't matter as much, right? onshape has a cyclodial tool for free, it looks liek fusion 60 has one that's $20 probably several videos worth of ideas here :-)
Yeah it often happens like that. Lots of rabbit holes to explore, I have a rather long list that fills up with topics much faster than I can check them off.
@@antalz As for perks you might make source files of what you model available for patrons. For example, your explanation of bevel gears modeling is really great and I understood how it's done, but nonetheless, I lack skills and time to model it properly for my own hobby project. And in such case patreon allows me to thank you in some material way and unblock my project, and for you it won't be like throwing your efforts for nothing. Win-win:)
@@gppl77 I currently usually just make the files available for free, I just sometimes neglect doing so or I feel the files don't have much value in themselves. Whatever the case, I would prefer as much as possible to be available to as many people as possible. I know there are also viewers from Brazil, Egypt, India, Sri Lanka and many other places that I would prefer can look at the content without paywall. I'll give it more thought, but I have a fulltime commitment now that'll last until March I would say. That's also why I haven't been posting anything unfortunately.
Great video. Based on my reading, the first right angle line is 1.25M and the second is just M. I believe this represents the total tooth depth. Let me know your thoughts.
Yeah it's like with racks, and also regular gears. From the pitch line/circle inwards is the dedendum, which is 1.25*module standard, and outwards is the addendum, which is standard 1*module. It's a bit clearer in my video on racks.
@@antalz This may not be critical but you made yours 1.25 * module on both sides in your tutorial. Great presentation. I love how you explained it all. I am new to CAD and I decided to learn Freecad only because it will remain free. From all that I have seen on UA-cam, Fusion 360 seems to be better but it may not remain free for long.
@@dannyperry8070 Yeah I did. That's because I want to create two cones, for two gears, and later add teeth to them. The "inner diameter" for both cones is their pitch diameter minus 1.25*module. When measured from the pitch diameter, the tooth goes 1.25*module down to meet up with the body of the gear, and 1*module up. This causes a 0.25*module gap between the tip of the tooth of one gear and the inner diameter of the mating gear. This is intentional clearance.
@@antalz Thank you for your quick response. So when I draw my sketch in Freecad, I should use the 1.25M on both sides? I don't know if it makes much of a difference in 3d printing but for accuracy's sake? So the .25*M is just and allowable clearance used in the industry?
@@dannyperry8070 Yeah for making the "inner diameter"-equivalent cones, definitely use 1.25*module on both sides. It's actually more important in 3D printing, because the process isn't that accurate, so clearances are more important, not less so. If you don't have this clearance you risk the tooth tips interfering with the other gear. I'd recommend modelling up two basic spur gears and looking at their dimensions closely. Check what the pitch diameter, inner/outer diameter, addendum/dedendum are, and look for that clearance near the tooth tip. I think it's wise here to understand that first before continuing. As you can tell from the length of the video, bevel gears are somewhat advanced stuff, and even this 40-min video builds on stuff done in previous videos. I fear you might get stuck later otherwise.
Thank you so very much for this. My perfectionism did not allow me to settle with any of the countless videos that show how to model "gear-like objects" that dominate the search results when trying to learn. This tutorial was perfect. A proper approximation, brilliantly explained and demonstrated. I have learned so much from this and will be referencing this series quite a lot, I'm sure. Thanks again!
Thank you! This episode especially was a bit special to make in a way, because the first videos in this series were mostly made because existing tutorials were usually incomplete. For this one though, most other tutorials are just plainly wrong.
I realise you published this ages ago, but still wanted you to know. between this and your worm gear video (the only two I have watched so far), I have learnt far more about fusion 360 than I have in the past 5 years of hobby use. That's despite watching a number of other tutorials and countless hours or googling.
Thanks, that's still encouraging to hear!
I've also just come across this video, and now I intend to devour all the other videos of yours.
I have to add my voice to the choir: Simply outstanding! Honestly one of the most educational, interesting, and USEFUL series of videos I've seen on UA-cam. THANK YOU for the thoroughness and fantastic presentation. I'm just sad it's taken me so long to discover these!
One of the best videos i've seen on modelling bevel gears. Keep up the good work!
Thanks a lot, and I'll definitely try my best
This video saved my butt. I was trying for days to model a bevel gear set for a project, and the mesh never worked out. I didn't understand about the back cone and virtual gear. Your explanation was perfect. I am using SolidWorks, its wild how different F360 is.
The thorough explanations are very helpful. Thank you so much!
This is the best video out there for understanding and modeling bevel gears. Thank you!!
Thanks mate, and good luck with your project
One of the best videos I have ever seen - increadibly excellent explanation!
Thanks mate, you're making me blush
Just now seeing this. You did a fantastic job. I don't think I've ever heard such a clear explanation of a mechanical system.
Here from reddit. Great video!
Welcome, and thanks!
What is modules?
When I draw the pitch circle line I get a dimension of 80.82 instead of your 89.84. All the rest of the drawing looks correct. What am I missing?
Cant you put a bearing and support in front of the large gear before the shaft of the small one to stop the cantilever?
Yeah that's definitely possible, but it can be tricky to put the bearing there and still make the whole thing assemblable. For high torque or high power applications it's worth investigating.
With two parallel lines selected, how did you create a constraint and then add the distance between them?
You can first push the button for the parallel constraint. It may not be on your ribbon bar, but if you click on the word "constraints" you get a dropdown from which you can select it. Afterwards if you select both lines and use the dimension tool, it will set the distance between the two lines. Usually it will set the angle, but it will not if you've already defined the lines to be parallel. I almost always use the shortcut D to use the dimension tool.
@@antalz, thank you
Very clear and complete instructions, as always!
But far more complicated as plain spur gears, and placement of the gears on the axes is quite critical as you explained.
I want to 3D print pretty small gears for use on perpendicular axes; about 10 mm diameter for the pinion to mount on the axis of a small Mabuchi type motor (as the ones in servo's), driving a large gear of about 80 mm diameter. After seeing your excellent tutorial, I'm now considering to make thinks easy for myself and use a regular spur gear for the pinion and a crown gear for the large gear.
But crown gears don't seem to be very common, except for in toys, tower clocks and wind mills .... How do they compare to bevel gears?
I'm still a bit confused by a crown gear to spur gear mesh. The way I understand it now is that it can only work with copious backlash and some friction. I probably just don't understand it correctly. The wind mills of old didn't have the technology for proper gears of any kind, let alone bevel gears. Modern turbines don't use crown gears as far as I know.
I think your problem can be done if the torques aren't too large, by SLA 3D printing. The pinion will be very hard to do with FDM I would imagine.
@@antalz Thanks for your prompt reply. I'll do some experimenting.
I just found 2 articles on this subject, you may find these interesting too:
www.geartechnology.com/issues/0901x/Weck.pdf
www.powertransmission.com/issues/1013/crown_gears.pdf
@@janpoppeliers8619 I'll take a good look at these, if I figure it out I have another video to make
Would you be able to do a video about helical bevel gears please? Also shoutout to the Gleason system ;)
I'm afraid not, I think I came up with a modelling method at some point, but I was never confident it was correct. UA-cam is full of incorect gear modelling videos as is. I also forgot what that method was exactly.
@@antalz I really appreciate all the work you've done on these videos, what you have to say about there being too many incorrect videos is unfortunately true. Thank you so much for all your help and guidance!
Awesome video!
Awesome content, now I have no reason at all anymore to further delay my project:-)
Great, good luck with your project!
What do you think would be the biggest load of a linear actuator propelled with 3D printed gears and what would be the best gear layout for durability. Thanks
My force-extension machine can produce about 300kg of force, but most of the heavy lifting comes from the M12 threaded rods and nuts. Still, there's no real upper limit if you make the gear large enough.
@@antalz Thanks for that Antalz. Would a proper ball screw be better than buying an M12 rod from a hardware store. For your project, would a M14 bolt work better ? If you were to use your machine to move a 50 KG load in the vertical plane, with a stroke of , say 15 cm, and a period of , say 5 seconds, ( 3 cm a second ) , what would be the best combination of threaded rod, plus gears. 3 cms a second may be a bit quick for a normal M12 threaded rod. Sorry, I don't wish to be setting you a task here. May I re-phrase the question to a simple yes or no. WOULD IT BE POSSIBLE USING 3 D PRINTED GEARS AND AN 895 MOTOR TO PRODUCE A LINEAR ACTUATOR THAT CAN MOVE A 50 KG LOAD OVER A 15 CM STROKE, IN A PERIOD OF 5 SECOND ? A simple yes or no answer would help, thanks. very much.
@@jackspratt4343 Yeah that should be possible. What you're asking is 500N at 0.03 m/s, which makes for an output power of 15W which most 895 motors can easily put out.
I'm not sure what setup you should use though. Probably a very low or no gear reduction on the motor. If you take an M8 rod, with a 1.25mm pitch, you need to rotate that 24 times per second which is a bit much. You can use bigger rods, because they have coarser pitch. You can also use TR8x8 trapezoidal screws, I'm quite sure they can take that load. For a TR8x8 you'd only need about 4 rotations per second, so that would require some gearing. I recommend buying a motor with integrated planetary gearbox in that case. It's probably better to couple the motor to the rod directly using a flex coupler. Do use the ones with orange plastic spiders, not a fully aluminium one. No need for 3D printed gears there. It's also very hard to attach a gear securely to TR8x8 rod, or M8, it's much easier for bigger rod like M12 and larger.
Basically, find a threaded rod that has coarse enough pitch that you can obtain the desired speed.
For my own device, an M14 rod has coarser pitch, which means my machine would move faster, but exert less force. Even low-grade M12 rods can carry several tons of load, so that's not a concern.
@@antalz That's great advice, Antalz. Thank you, May I ask whether the power calculation, of 15 Watts for 500 N @ 0.03 m/s would "extrapolate" by a crude rule of thumb, to 30 Watts @ 0.06 m/s and 45 Watts, in an ideal world @ 0.09 m/s.? Thanks again.
@@jackspratt4343 Yeah, energy is force*distance, and power is force*speed.
hello sir, sir can I ask what 3d printer you used?
I think the black housing came from a CR-10S4, the older version, and the other bits from an Ender3 and CR-20.
@@antalz copycopy sir
W video
You are amazing Man! I have learned a lot of valuable information from you. Thanks! Would you be interested in working on a small project with me? Thx
Thanks mate! I'm curious to know what kind of project you've got in mind. Could you send details to antal@creonova.nl?
Thank you so much for making these videos!
This is only of the only places on the internet to find all the correct info in a very well explained and complete manner.
You deserve so many more subscribers. I'll do my best to share your videos.
Thank you, and thanks for your support
Your tutorials are perfect. Thank you very much! I am building a 3D printed 6 axis robot arm and these videos are helping me to build actuators
I've been waiting for this one and needless to say it's a masterpiece! Your videos always feel like they are only 15 minutes long that's how gripping they are. btw your insta link in the description is broken :)
Thanks a lot, and thanks for pointing out the link, there was a trailing space there. Should be fixed now.
Fantastic video, I am learning a lot from this series. I would love to see your approach to making a helical bevel gear in Fusion. I'm not quite sure how to get the appropriate twist in the lofted teeth, and how to get them to mesh properly. That would be an interesting video.
Thank you! I thought of doing it using a loft with guiderails, but I haven't actually done it. I think you will also need to emboss the curve onto the pitch cone, but again still need to look into that.
@@antalz I tried with guiderails with both Sweep and Loft features, but I couldn't obtain a satisfying enough result. So I ended slicing a straight tooth, rotating each slice as needed, then executing the Loft feature with all slices as profiles.
@@steelstone That's unfortunate, that sounds like a lot more work than embossing a curve and then doing a guiderail loft.
Very nice work !
And I learnt something, we can make joints between components and sketches. Thanks !
Its things like this that we take for granted till we have to do it and then you realize the immense amount of math and things you have to wrap your head round that goes into projects/products we see every day, it just blows me away. I never knew there was so much to learn about gears, it always seemed straight forward to me till now. Thank you for this video!
Thank you for making this video. You are clear, concise, and thorough in your presentation. Exactly what I was looking for.
Hi! This video is crazy, it literally teaches very good, but i still have a problem unfortunately… I don’t really get how to calculate the dimensions that i must input to work well but to be much smaller than what you did. I know u published a very long time ago but i hope you will have time to write a quick explanation! Thank you:)
Thank you so much for making this, I really enjoyed how well explained everything was.
Incredibly well explained, Thank you, I know I will need this one day!
Thanks a lot, next part should go public within an hour or so. Forgot to rotate some scenes so rerendering now.
I don't like leaving warnings in my timeline. And whatever I design is nowhere near as meticulous as this. Leaving a warning in the timeline on your design would be like inventing a perpetual motion machine and not starting it. Still trying to wrap my head around this one as I am going to need it. Will need to do some dry runs before incorporating into the main design. Great job explaining it. I don't think I can find anything remotely close elsewhere.
Thank you very much for this video and in general for all the others. I knew this construction method mechanically but now I know the reason for it. In addition, the way to apply it in F360 is brilliant.
Thanks for your kind words, and thanks for watching
TOP, as always! Waiting worm gears with anticipation!!
Thanks a lot, hope to see you then
This is excellent - I used your method with Solidworks (and manually modelled involutes for the fun of it) and it works very well. Thank you :)
I must also say, that I am envious of Fusion 360 being able to animate parts and examine motion without having to swith to an assembly file. That is very cool...
Okay, I spent a couple hours trying to take a similar approach in Missler Topsolid, where I don't have a gear generator to lean on. It does have an involute curve generator though. I got to the point where equal teeth on the 2 gears mesh perfectly, but once I shrink the pinion, the teeth no longer mesh. If you have a few minutes to look at this video, that would be a huge help!
ua-cam.com/video/6ZFHzi5EphM/v-deo.html
hello,this kind of drawing won't make the bevel gears conjugate, as you say approximation. and we can make the real conjugate bevel gears,but not your method.
Do you know a modelling method for true conjugates? I don't think it matters much because the approximation has smaller errors than almost anyone can reproduce on their printers. But it'd still be interesting.
As far as I know even commercial producers use one of a number of approximations to ensure they can mill the gears.
Perfect Job!
Excelent video thanks. If you hollow the large gear it might just be possible to mount another bearing inside it between the gear and the shaft.
That's actually a very interesting idea, though it does appear to me assembly could get tricky.
Very interesting and the presentation of a difficult design was well done. Keep up the great work and I look forward to more of your videos.
I have a video ready for filming, but I still haven't done it. Hoping to get there soon!
@@antalz Please keep me informed,I am looking forward to viewing your latest work
Brilliant!
At 4:40 - You mention the line is "1.25 times the mod" , in this case you entered "2" for the mod. what is the mod in this case? It seems like it is the tooth depth, and if so, why not call it that instead of a "mod" ?
The module is a measure for how large the gear's teeth are, and is related to the pitch diameter. In this case the bevel gears we're building are module 2 mm. Now when measured from the reference cone (compare reference circle or pitch circle for normal gears) I want the teeth to have an addendum of 1*module, and a dedendum of 1.25* module. The gear generator will take care of the addenda, but for the dedenda I must specify here that the cone that makes up the body of the gear is 1.25*module smaller than the reference cone.
For bevel gears you essentially do all the math for the outside edges of the cones, and then from there loft everything inwards to the apex. This also means the teeth are larger on the outside than on the inside, so when we say "module" for these teeth we mean the module on the outsidemost end of the tooth.
goat
could we theoretically draw the 2d face for a regular gear, and do a loft to the centre point, then "cut" the body?
Yeah that should be possible too, it might even be better because you can get that fractional virtual number of teeth. Drawing an involute manually is a chore though.
I am not the person who is making comments etc. But you really deserve Bold and Huge Thanks, my friend. This video and the thing you shared is tremendous. It has never been easy this much. Thanks for being clear, and helpful, and for your greatness. I am a huge fan of you after all!
Superb presentation - thanks for capturing and sharing this very clever practical technique so professionally and concisely.
The best making gears!! Your channel just earned a new subscriber.
This is great but I tried it for a 30mm diameter, so radius would be 15. Approximation is around 21.25 so I take it 22, but the gear tooth stays to outside for module 2 gears. 21 also stays to outside...
Do you mean you're getting interference? This might be because if you're doing 15 teeth undercut is required, which the Fusion360 gear generator doesn't do correctly. Even then undercut is better avoided if possible, a larger pressure angle may help.
@@antalz Hi there, thanks for the answer, I will try different pressure angles.
No, I meant when you align the single tooth with the plane you created along the intersaction axis, your gear tooth is almost halfway embedded inside the cylindirical body. In my case its not embedded, kind of touching in bottom middle, but edges are lifted.
Maybe it is normal and I need to sketch and extend so I can create the solid feature, but it feels like it is wrong and I might have problems later on.
@@ROBOROBOROBOROBO If I remember correctly, the tooth getting properly embedded in its gear occurs one or two steps later when the base of the tooth is push-pulled a bit to embed it properly.
This is the best video about bevel gears! But could you also make one about spiral bevel gears?
At the moment no, I don't know for sure how to do it. I had a method in mind, but I'm still not sure whether it's legitimate. UA-camr "steel stone" has a video: ua-cam.com/video/L_hcaxxKt0g/v-deo.html
I am getting small amounts of overlap on the meshing of the teeth when I make gears with a higher module even with backlash, I do everything else the same except in the case of the gears I am trying to model they are a ring and pinion with 43/13 ratio and a module of 3.5 and I am using a backlash of 0.1, should I just move the gears further apart to force separation like in your pt. 2 of the gears series?
If the overlap is small I suspect it's because of the 13 tooth gear. That is below the undercut-free limit of I believe 17 teeth for 20 degree pressure angle gears. Either increase that tooth count, or a pressure angle of 25 degrees should work. Fusion's generator can not model undercuts. Also check you have a correct combination of gear ratio, shaft angle and cone angles.
@@antalz oooh thank you very much! I suspected that I should increase the pressure angle as it's a high torque application, this model is for a transfer case. The overlap is very small but visible by eye so it would bind in a 3d print for sure, maybe I could manually add a small undercut for low pressure angle gears with a small number of teeth? I believe my combination of ratio, shaft angle and cone angles is all good, my sketch looks to be correct and I used simple formulas like you did but with my 3.5 modulus in place, the joints and motion links also work correctly minus the small overlap which is consistent on every tooth so positioning seems correct. I'll try the higher pressure angle and if that doesn't work I'll manually undercut the teeth, thank you for your help!
@@antalz After a bit of investigation I found that even with a pressure angle of 25 degrees I had a slight overlap of the teeth, when I had the backlash set to 0.1mm I found that moving the pinion 0.4mm further away from the ring gear allowed the teeth to mesh well but technically now the pinion would be off center of the correct position of the shaft, would this have a significant impact on the performance of the gears?
@@liamgleason3416 I don't think that'd have significant impact for M3.5 gears. Still strange.
@@antalz thanks for your help! I always go back through your videos when I'm modelling gears to double check my work, they're the best!
Hi, great video
I'm interested, what would happen if instead of using the gear generator, you make the tooth profile for the gear with 25.73 teeth or any non integer number.
Would it be reduce backlash between teeth, or it would cause them to interfere?
Thanks for your time
I think you could in theory make a slightly better fit with that, but the profile difference between 25 and 26 teeth is very small. Very few people are making bevel gears at such precision that it matters.
Tx for sharing that completed gears lesson. Btw, you told to manually adjusting the layers height within the cura's slicer, with bigger at base, thinner at top ... but how do we do that ?
In Cura you should be able to get started by ticking "use adaptive layers" under the experimental features.
I still get confused when to be working in the original part or the separate components. When the gear tooth is lofted and trimmed, you have the bullgear active, but when the teeth are patterned, you go back to the original part to pattern them. Why aren't they patterned with the bullgear active?
I went back to the root component there so I could edit the big initial sketch. The thing is, if you do an operation for "new body" or "new component", those get put into the component that's active at that time. So for the initial loft I'm doing a new body, so the bull gear must be active so that body ends up in the bull gear component. For cutting and patterning it doesn't matter, those features end up in whatever component you were cutting, or in the component that held the thing you were patterning. In fact a cut can affect multiple components. If you look on the timeline you see that the pattern and cut both ended up with the color of the bull gear component.
hey Antalz I love your tutorials they are very informative! I was wondering if it wouldn't be too much trouble could you further expand on this tutorial by adding how you could create a helical bevel gear?
Thanks mate! At the moment I don't quite have time to make videos unfortunately, but I'm hopeful I can get back into it. The bigger problem is that I don't know how to create a helical bevel gear. I have some ideas but I don't know if they create correct gears.
@@antalz I'm still playing around with it but that's my problem as well. I'm thinking that i need to use the contact lines length to drive the twist angle. I then made a offset plane at 50% of that and again at 50% of that as well and then divided the twist angle accordingly and rotated the original loft sections by those angles and created a new loft
Are there any reasons to not have the contact lines/inner diameters of the two cones parallel to each other? Wouldn’t that offer more contact area? I may be wrong, just wondering.
No that just doesn't work geometrically. As you get closer to the apex, the gear ratio and number of teeth have to stay the same. To make that happen, the teeth have to become smaller as you get closer to the apex. For that to happen the inner diameters must converge onto the apex point, meaning they can not be parallel.
Maybe you mean making the teeth longer, but that doesn't actually increase contact area; it just makes the teeth more likely to break off.
very smooth.
Thank you very much! I have struggled over a week to make bevel gears in Catia and your video helped me get the last piece of the puzzle
Great to hear, thanks for watching
Thanks a lot man you rocked my design work....
Thank you, and good luck with your projects!
This is a good approximation. The spur gear used to give the tooth profile is normal (at right angles) to the pitch cones line of contact. I've seen where others set the spur gear rotation axis equal to the bevel gear rotation axis. This causes loss of tooth depth but not tooth profile width and therefore a very distorted bevel gear form.
Yeah that's a big reason why I made this video in particular. There are a lot of videos out there on gears that are wrong, but specifically on bevel gears there is a lot of misinformation.
@@antalz Thanks. I've been producing gears for decades on my CNC lathe and milling machines. Only ever spur gears. Now with 3D printing, I'm getting into other applications and one job requires bevel gears. So where do you go to learn such design? LOL. I love UA-cam but as you say, there's plenty of bullshit out there. I'm an Aussie. We don't say "misinformation" That's way too polite. LOL.
really good video, thx for your time to save mine :)
You're very welcome, thanks for watching
Muito bom amigo!
Very awesome and informative! Thanks for sharing!
Thanks a lot!
Would you ever consider doing a video on spiral bevel gears?
Yeah, but at least for now I'm still not sure how to do it. I have some idea but I'm not sure if they result in a correct gear.
@@antalz understandable; amazing videos in the whole series anyways. I have little note cards sitting with the formulas you explained. Put a lot of perspective on the caveats and variables of desighning gears.
Waw so detailed! Great stuff!
Thank you!
what was that, in between the bull gear and the motor, which drives the bull gear, attached to the motor(29.30)?
Not sure what you mean. The blue gear is attached directly to D-shaft of the motor. Between the pink gear and the bearing I put a small 3D printed spacer.
@@antalz i mean to say, the motor you had used has shaft diameter 5mm but in the gear hole is of 8mm...that black part top of the motor is extra part? To increase the torque or it is the part of motor? Or can you please provide another link where i can find the same motor ...i don't know but that link has some trouble 🥲
@@kamrealamshaikh4375 Yeah the 775 motor itself has a 5mm output shaft, but the motor I bought comes with a planetary gearbox included and attached. I think I used the 24V i13 version here: nl.aliexpress.com/item/32822227433.html
I used the same motors for my tank project, which I still need to finish.
@@antalz yups, that's what i want to know....Thanks, great work by the way. I am using your design in my project that's why I need to know about that gear system 😉😉😉
Excellent tutorial. One quibble, though. At 37:30, if the tooth profile is the (almost universally-used) involute form then, even though the axes are displaced, the teeth still roll against each other without sliding. It's just that the pressure angle is increased.
I didn't know that, thanks. Do you have any source where I can read more on this? Does this relate to profile shifting, because that's another thing I should learn more about.
@@antalz The fact that I mentioned is at 4mechtech.blogspot.com/2015/06/comparison-between-involute-and-cycloidal-gears.html. If the gear teeth are too narrow at the base then increasing the pressure angle makes them wider at the base, at the expense of an increase in separating force along the line joining the gear centres. If you Google "involute profile shifting" and choose the Marples Gears site there is more info on profile shifting. However, I still don't understand how a profile-shifted gear meshes correctly (i.e. rolls, not slides) with a non-profile-shifted gear.
@@chrisarmstrong8198 I'm not sure that site says the contact remains rolling. It does say the velocity ratio remains constant, and I think that's correct. I also think you're correct that the pressure angle increases because a further-out piece of the involute contacts. I'm just not entirely convinced on rolling contact, though to be fair it took me a while to see it's perfectly rolling for a perfect involute set. I'll study this some more, thanks again.
@@antalz I offer this hand-waving justification :) Consider two rotating cylinders, with parallel axes, joined by a piece of taut string marked with equally-spaced dots and arranged so that, as the string winds onto one cylinder, it unwinds from the other. The dots trace out the involute flanks of (one side of) imaginary gear teeth on BOTH cylinders, with the dots themselves being the contact points. (To trace out the other half of the flanks, the string needs to be wound in the opposite direction on each cylinder.) The relative velocity of the contact points on two mating imaginary gear teeth must be zero since they are produced by the same dot. Changing the separation of the cylinder axes does not change this fact. Zero relative velocity at the contact point implies rolling contact.
@@chrisarmstrong8198 Hand-waving maybe, but convincing! Thanks!
Thanks!
My pleasure, thank you!
Helpful.
Thanks, that's good to hear
wonderful series (I just stumbled across it), I would love to see an entry on cyclodial gears as well (given that in the planetary gear episode you talked about doing compound gears for very high gear ratios), both single stage and double stage (which eliminate the pins and bearings for the output shaft)
do you have a patreon account or similar to allow us to support your efforts?
Cycloidal drives is actually a good idea, I have seen a number of videos of people who made them on UA-cam, the latest being James Bruton. However I do believe most of them don't talk in-depth about how they designed it, or what things to consider.
I haven't yet opened up a Patreon or anything like it, but it's something I'm starting to consider. Not entirely sure what to do with the levels and perks though. AvE seems to do quite well with his single-tier beer fund, but most creators have some progression of tiers and perks.
@@antalz re:patreon, I sponsor a fair number of people and mostly ignore the perks, start off with a beer fund approach and if it turns into real money, you can add perks that make sense based on requests (you have a business, so any perks you offer need to make sense from a business point of view) Applied Science does per-video sponsorship so he doesn't feel pressured to get a video out on a particular schedule, which is good for in-depth technical videos.
re: cyclodial gears, yes I watch a lot of videos from people who are making them, but other than the gear ratio, very few people talk about the considerations in designing them (especially for 3d printing)
what is the real effect of different eccentric values (is it like pressure angle)?
I see some people with bearings on the outer pins, others without and talk about it being a rolling contact, what is real?
almost all cyclodial gear setups are N:N-1 tooth ratios, what if you did something different?
dynamic balance of cyclodial gears is an interesting topic, but I think it only matters if you are talking about higher speeds where the disks are heavier. If you talk about slow things (camera mounts, 4th axis for CNC, etc) this doesn't matter as much, right?
onshape has a cyclodial tool for free, it looks liek fusion 60 has one that's $20
probably several videos worth of ideas here :-)
Yeah it often happens like that. Lots of rabbit holes to explore, I have a rather long list that fills up with topics much faster than I can check them off.
@@antalz As for perks you might make source files of what you model available for patrons. For example, your explanation of bevel gears modeling is really great and I understood how it's done, but nonetheless, I lack skills and time to model it properly for my own hobby project. And in such case patreon allows me to thank you in some material way and unblock my project, and for you it won't be like throwing your efforts for nothing. Win-win:)
@@gppl77 I currently usually just make the files available for free, I just sometimes neglect doing so or I feel the files don't have much value in themselves. Whatever the case, I would prefer as much as possible to be available to as many people as possible. I know there are also viewers from Brazil, Egypt, India, Sri Lanka and many other places that I would prefer can look at the content without paywall.
I'll give it more thought, but I have a fulltime commitment now that'll last until March I would say. That's also why I haven't been posting anything unfortunately.
Great video. Based on my reading, the first right angle line is 1.25M and the second is just M. I believe this represents the total tooth depth. Let me know your thoughts.
Yeah it's like with racks, and also regular gears. From the pitch line/circle inwards is the dedendum, which is 1.25*module standard, and outwards is the addendum, which is standard 1*module. It's a bit clearer in my video on racks.
@@antalz This may not be critical but you made yours 1.25 * module on both sides in your tutorial. Great presentation. I love how you explained it all. I am new to CAD and I decided to learn Freecad only because it will remain free. From all that I have seen on UA-cam, Fusion 360 seems to be better but it may not remain free for long.
@@dannyperry8070 Yeah I did. That's because I want to create two cones, for two gears, and later add teeth to them. The "inner diameter" for both cones is their pitch diameter minus 1.25*module. When measured from the pitch diameter, the tooth goes 1.25*module down to meet up with the body of the gear, and 1*module up. This causes a 0.25*module gap between the tip of the tooth of one gear and the inner diameter of the mating gear. This is intentional clearance.
@@antalz Thank you for your quick response. So when I draw my sketch in Freecad, I should use the 1.25M on both sides? I don't know if it makes much of a difference in 3d printing but for accuracy's sake? So the .25*M is just and allowable clearance used in the industry?
@@dannyperry8070 Yeah for making the "inner diameter"-equivalent cones, definitely use 1.25*module on both sides. It's actually more important in 3D printing, because the process isn't that accurate, so clearances are more important, not less so. If you don't have this clearance you risk the tooth tips interfering with the other gear.
I'd recommend modelling up two basic spur gears and looking at their dimensions closely. Check what the pitch diameter, inner/outer diameter, addendum/dedendum are, and look for that clearance near the tooth tip. I think it's wise here to understand that first before continuing. As you can tell from the length of the video, bevel gears are somewhat advanced stuff, and even this 40-min video builds on stuff done in previous videos. I fear you might get stuck later otherwise.