home made epoxy granite cnc - part 2 finishing the base
Вставка
- Опубліковано 29 бер 2019
- Second part of my epoxy granite cnc build.
First getting the aluminum supports as straight as possible on the mill of a friend, and then trying to assemble the linear rails as accurate as I could do within my possibilities.
Thanks a lot to Andy of Desfiladero / @desfiladero for the great milling job!
A bit more details on this build in written form you'll find here: dangeroustools.com/category/cnc/
Part 1: • home made epoxy granit... - Наука та технологія
1:25 That mysterious reason is that when you clamp the part on the milling machine, you apply uneven pressure on different regions of the work piece
I was about to comment the same 👍
Nice to see you again we were waiting after our lesson that steel webb welded construction techniques. Looking great will be here for the part 3. Lance & Patrick.
Without precise measuring instrument, flipping the measuring tool method is a clever way to do。hope to see the coming video。
Very nice video showing the metrology reverse method - that's cool! Just one remark/word of caution: Measuring straightness of the linear rail itself is dangerous. You are assuming the grind on the rail is perfectly parallel to the line you are measuring on the rail. That's not always the case with cheaper linear rails! I saw more than 10 microns deviation on 0.5 meter...
Better is to measure against the real carriage sliding on the measured rail. That way you remove one more unknown. ;-)
Great work! I'm fairly certain you can get a rail straight by using one indicator mounted to a linear bearing, slide it down the rail, and adjust until within tolerance. Then use the first rail as a reference for the second one. Personally I just bought a 24" granite straightedge.
Really nice work.
great job so much effort
Cool work
Well done and excellent music selection ,-)
The point between your cutter and the table will always be the same distance because the head doesn't move and the bearing blocks always come to the same spot. so i would check the surface after it is all together before i got to worried. I would check it as a whole unit and then adjust the rails with shim stock or use a dial indicator from the head that is stationary cutter doing the cutting work. I have had to setup old mills that had alot of wearing space in the bearings and the cutter would vibrate so i shimmed up the cutter inserts to the bed starting with the tooth closest to the bed, i will tell you it made a differences of no shine rough finish to a finish you could see your face reflection in. There alot of GM cars running down the road with those V-6 water pumps on them.
Pragmatic choice to get them done in a machine shop. I really do want to try doing by hand though.
You should definitely download the datasheet of the linear rails and find out which surfaces are specified for tolerance. For example, even though it looks precision ground, the top surface rarely if ever has a tolerance specification. Also, indicate the distance at the base instead of the ball races.
Mohammad was mostly correct. You should have shimmed the bottom of the part before milling. There was almost certainly a bind in the part from clamping.
Easier way is when your milled the rail mounting faces, mill a reference surface along the side. Then you just indicate everything from that.
Yes, if the milled surface would be perfect. The top surface wasn't, had to put a few shims in. Still not sure why. Guess it's because I used only four feet (m10) instead of six to level the piece in the milling table. And 10mm feet might be a little weak as well...
@@joergbeigang because the mill table isnt leveled...
@@spikeypineapple552 better yet, get it surface ground. for final measurement and perfect flatness, nothing is better than a surface grind. Just ask your local machine shop to do it. A lot, seeing you are making a machine base may only charge you cost, the wheel's are a bit pricy, but very little labour is involved. However, they will probably want to talk shop with you for a long time :D
@@Keri-Kerigan Lapping and/or scraping is better than surface grinding. But that's indeed practical only if you need accuracy in 1 micrometer range or the work piece is too sensitive to even mild clamping forces (like during grinding)...;-)
I wonder if it is possible to make a surface grinder by this method? Thank you.
Really enjoyed this series. Hope it’s working well for you. Would you have a link for the tracks and runners you used? I looked around but hard to judge quality etc online.
These are CHTR tracks. I'm not experienced with linear rails, it's my first cnc build, so I can't really compare them. Till now they are doing their job, that's all I can say.
Joerg Beigang Cool. Thanks for getting back to me. Hope it stays working for you 👍
nice idea
I recommend using the X axis of your mill for the first rail, then line up the second off the first. I like your use of the reversal method but the wood has to much error despite a good approach. The project is coming along nicely!
Thanks for your feedback. I know the wooden reference doesn't look very trustworthy. In the video I just showed enough to tell the story. In fact I did several measurements like this, also because I had a bug in the spreadsheet, and compared the results. Got to say for me it worked. There's of course no comparison with using a proper reference (which I don't have). And that's the thing I like about this reversal thing, the reference doesn't need to be straight, it just got to be the same on both measures. Got to say it's a nice journey for me, I'm learning a lot doing this.
It may be possible. Clamping the straight edge to the wood, then indicating from the first rail truck might have been easier and more precise. Since desired results is the truck to run true. But, the trick is setting the reference plane without a surface plate, gauge blocks or precision level.
@@joergbeigang Yes, a wooden bar as a "staight" line or even a support for a non-straight rail would I use as the last thing, actually never. You have a nice milling machine, why don't you dial on that, it's straight. An other tip is the second rail you can self align with the mounted X plate easier and more accurately.
@@lp7489 the funny thing about the reversal method is that the reference doesn't need to be straight. It's not getting more precise if the reference is straighter. I decided to go for a piece of wood as a proof of concept (if it really works its gotta work with a pice of wood as well). The only thing the reference got to do is not changing it's shape during measuring, and it's got to be measured at exactly the same points. On the reference and on the target.
The milling machine isn't mine. It's in a friend's shop five hours drive from my home.
i dno't get it at all but you seem to know what you've been doing and the outcome looks fine
Thermal expansion problems are to be considered. All polymers and cements shrink. The aluminum could be distorted by this polymer cement base because both have a different thermal expansion. In ceramics, when you apply glaze onto a ceramic coffee cup to do a glass finish, the glaze must have an equal thermal expansion to the clay ceramic base or it will crack from mechanical stresses. Also when your milling the aluminum, it will expand a lot more with heat from cnc friction and will distort because it is held by the base which is not expanding equally. I think that is also a problem for high precision on low cost cnc machines using both plastic and aluminum for their structure. Still your idea is interesting but everything must be one type of material.
Haven't been able to find it in the comments or videos, yet (might have overlooked it), but what size rails are you using? 25mm? And how about the ballscrews and their pitch?
The rails are 20mm and the ball screws are 16/5 mm.
@@joergbeigang Thanks so much:). Great build and super inspirational.
Gues the clamping down to the table introduce a bend in the middle so it took a bit more from the center of your base.
The right way was to clamp it somehow to the mill bed with the srews you have under your base.
Just a though though, i really must be wrong.
:O
Brilliant!
nice build, never had seen epoxy granite before. What are working dimensions going to be?
travel will be 320 x 240 x 180 mm
Light Machines Prolight and Benchman CNC Machines used a granite co-polymer base.
thanks!
Hi. How much time passed after curing before machining? and why you don't made reference edge when you made machining?
I second the question about the curing time :) Great work, by the way.
The milling happened about one month after casting. But that's because of logistic reasons (the shop with the mill is five hours drive from my house, had to find a date that worked for both of us...).
The resin was hard enough to get it out of the mold after 24h. Ideally I'd wait a few days to a week to do the milling.
@@joergbeigang casting should be cured about 1-2weeks. And made referencr edge if you can, only 2-3 mm deep, but its make your life easier and mashine will be more rigid. Great project.
Nice build!
Btw. what's the first song?
Thanks. It's "Lenox", an unreleased by Moby. You'll find it at mobygratis.com/
Why didn't you create a straight reference edge for the first rail?
Is there any reason why you first calibrated the mounting bars on the casting side rather than the proper side?
Because I predrilled the holes, and there was only very little play with the M6 screws in the holes of the rails, I just wanted to make sure, the mounting bars are parallel enough. Didn't want to end up widening the holes of the rails in order to get them parallel
you set it up poorly in the mill, that why it's not flat or straight (assuming the mill is true and straight too)... most of the time we set up rails, of any kind, we mill a small shelf for the rail to sit against, so you have a flat surface on the bottom, and a straight edge on the side. and if you need to shim it, just use very small piece of shim, and before your final shimming, lay some bedding compound under the rail (I use Devcon), put a thin layer down then lightly torque down the rail to the shims, check it for straightness, get it straight, then wait a couple of days, check it again, then torque it down fully.
So, now that you know how to,when are you going into production and when is it going to be available
Got no plans in this direction so far. Not even sure if there's a market for such a small machine.
If I chose to do it, I first had to think about reusable molds made of steel, which I can't produce by myself.
The way I was building this machine was far from efficient.
But could be an interesting project...
@@joergbeigang I am sure there's a market but even better for us DIYers would be if you would either share the 3D files or sell them at a reasonable price. I'd for one would be very interested. It's the perfect build volume for me:)
Old fashioned scraping with a large granite stone to get high spots would create a flat surface.
I am a novice here, so please forgive me if I am saying something stupid. It seems to me that the underside of the casting isn't flat. When you clamp it down on the milling table you have to be very careful not to wrap it a tiny amount. Otherwise, you will mill the top flat only to have it warp again when you release the clamps.
This is a normal question. In this case, more powerful fixing of the part is more important. And the fact that there is slightly uneven - you need to create a flat base and you always have to push off from something - in this case, from the uneven back surface. Then, if necessary, you can use a new clean surface as a base and grind the back for beauty.
Why you mill this surface? You get flatness from the granite plate from first video.
WOW SUPER BRO
probably not the case but if your rails are stainless steel don't mix them with aluminum, they react.
Love the porn music at the beginning, it really sets the mood.🤣
the way you clamped it
reason it would be out
Dissimilar materials, yikes! You have to machine and accurize at the same temperature at which you will be using the router. Many times I have demonstrated how an indicator reading can change significantly just by breathing on the measured part.
too much mistakes
lol enlighten me, name a few
@@hyperhektor7733 e.g top side of guide is not a base side but you compare it using ruler. surface can be bended(propeller effect) and you measuring differences between two not straight guide .
seems like you overthought this process and got a bit lost. if the top side where the rails were mounted was all milled down .020 ( flat ), then the 2 rails would be on the same exact plane regardless of the back of the casting;. Then the relationship would be between those 2 rails ( which would be true to themselves ) and the rest of the machine...partially milling the topside with no actual reference point was pointless. this old tony has a video just like this when he builds his gantry router.. he has to outsource the milling but does exactly as i describe for the gantry..
I can't follow you here totally. With milling the top surface I defined a plane (never mind the shimms here, they are only in the center). By aligning the first rail in x direction I defined a (hopefully straight) line on the plane. That's all the reference I need.
The same procedure is gonna happen with the base of the z axis.
Later I will have two planes with a line on. When assembling the two pieces I got to make sure that each line is normal to the other plane (y to z and vice versa).
I'm planning to do it like this because I can't make sure there will be anything perpendicular on the casts. To fix this I'll mount the z piece with a small gap to the base, between two metal pieces left and right. Once it's perpendicular in both directions I'll fill the gap with epoxy granite to create a form fit.
@@joergbeigang lets assume , when the structure is laying on its back the rails are facing the Z direction.. Why , if the surface underneath was completely level , were you checking to see if the rails were level in the Z axis ( i understand the x and y , and how thats complicated )).. Wouldnt they automatically be level based on the previous milling operation to flatten the base ?? In other words, were the rails on a surface plate and measured with height gauge , wed assume theyre perfect , right .. SO why would they no longer be perfectly level after being installed on your freshly milled surface ??
another way to say it , if the milling operation leveled the bed , then the rails are precision ground , why would they no longer be level on the bed ? anyway i like the project and will follow.
I see, only half of the story is told in the vid. I should work on that. I didn't came perfectly out of the mill. There was a little dip in the center. About 0.01 to 0.02 mm. Can only guess why. But anyway, it's only a bit in the center beginnings and ends were fine. We measured that. So what I did was putting shims in the center area to make it straight in z.
@@joergbeigang ok.. understand now.. the sole solution would have been to mill the entire surface till all areas had been milled..( maybe .020 thou deep) then it would be flat.. anyway best of luck ill be following