V2 & V3 Drafting project.

[zzzzsleepy8]

Originally Posted by Bandit50

If I could 8mm SKF bearings, there made in France , Not china. Sealed would be nice since airsoft guns do get wet, but that is almost impossible. So, ill go with Uk, or Element and considering the gear box I linked a page ago... I think it's pretty obvious what I am doing and that i know what I am doing.

However, you can keep telling me I don't, I like it

LoL. Not sure why you are on a bit of a power trip and being defensive. Pretty sure a public post is for public discussion and nobody is telling you you're doing things incorrectly. Its just a discussion, relax. If you want to be boss giving out all the lectures, you can explain so in the OP so we can crawl in the corner next time. Honestly I could care less how you do it or what you do, though it will be interesting to see progress you make for curiosity sake. You don't need to listen to anybody if you don't want to. LoL. Free country. Again, I could have the V2 gearbox completely CMM and CADed and maybe even machined in max 1-2 days myself if I cared enough to (which I completely don't), instead of talking about doing it for years on end without doing it.

I'm still pretty sure though, there is no concerning of 'gaps for grease' if one goes commercial bearings route though. Anyway, the weather is probably different at your level so good luck.

[Bandit50]

Originally Posted by zzzzsleepy8

Now does this project involve machining your own bearings too (perhaps with a dash of aluminum welding and tempering)? Btw, commercially available bearing parts from reputable manufacturers are fit to tolerance too.

Not sure what the weather is like at your level, but pretty sure I would personally just buy the bearings from a commercial source for what its worth. And then I'm pretty sure I will pick the correct style of bearings and choose whether or not I'd want it to be sealed and internally greased or not. I'd probably go with a sealed one. So again, don't see how greasing bearings becomes an issue I'd have to worry about, or the 'leaving of gaps' for the grease as far as commercially available bearings are concerned.

This project involves measuring, and drafting it, in CAD form from a factory spec, if you read the first post of this thread it explains that. This is to provide a plethora of drawing to the community for educational purposes.

At the same time, it's something I can experiment with in CAD, to show others that there is more than just one type of cookie cutter design, and if I find something more efficient, and direct. Ill be sure to inform everyone.

Originally Posted by zzzzsleepy8

LoL. Not sure why you are on a bit of a power trip and being defensive. Pretty sure a public post is for public discussion and nobody is telling you you're doing things incorrectly. Its just a discussion, relax. If you want to be boss giving out all the lectures, you can explain so in the OP so we can crawl in the corner next time. Honestly I could care less how you do it or what you do, though it will be interesting to see progress you make for curiosity sake. You don't need to listen to anybody if you don't want to. LoL. Free country. Again, I could have the V2 gearbox completely CMM and CADed and maybe even machined in max 1-2 days myself if I cared enough to (which I completely don't), instead of talking about doing it for years on end without doing it.

I'm still pretty sure though, there is no concerning of 'gaps for grease' if one goes commercial bearings route though. Anyway, the weather is probably different at your level so good luck.

Power trip? Is that why you keep trying to find the issues in everything I write and correcting them to make yourself feel bigger? >.> Yeah, I am going to be defensive, after you publicly ridicule me.

Also, have I been talking about this forever? Pretty sure I said last year the project will get started in spring, do you not have patients it is march after all? If I wanted to give out lectures, I would have made a youtube video and linked it. So yeah, this is a public discussion, but you know have some respect before you jump in and hijack it and say i am doing something, when I am not.

This was about drafting, and ideas. Not making a gear box from a ingot of steel, and using a CNC machine. >.> Pretty, sure I have said that like 8 times now.

Experimenting in CAD, is making something in a simulated 3 dimensional space. It isn't drawing something walking over to a machine uploading it through a USB and test firing after it's assembled. What do you think this is, 3d print a gun? Not what I am going for here. It's more or less concept that will be a series of screen shots very much like the 2 things I have already done. The only IRL/live shot you will see is the parts I ordered and the measuring process behind it.

[zzzzsleepy8]

Originally Posted by Bandit50

This project involves measuring, and drafting it, in CAD form from a factory spec, if you read the first post of this thread it explains that. This is to provide a plethora of drawing to the community for educational purposes.

At the same time, it's something I can experiment with in CAD, to show others that there is more than just one type of cookie cutter design, and if I find something more efficient, and direct. Ill be sure to inform everyone.

Power trip? Is that why you keep trying to find the issues in everything I write and correcting them to make yourself feel bigger? >.>

Nobody will give you engineering drafts or factory specs so you can go CAD it up. Good luck with that. I'm pretty sure most of the clones of TM AEGs in the post TM V2 gearbox era basically took a TM apart, reverse engineered it in way of CMM and CADing the parts, and then machined their own. The gearbox clamshells are pretty simple in design. It wouldn't take long to get a fairly well spec-ed clone CADed provided you have the ability and tools.

Again, I'm not the one on a power trip talking about 'your level' and 'that level' while ironically still using an educational version of autodesk inventor. Its unfortunate you feel so as everyone is just providing a discussion. In any case, I think its a neat hobby project anyway. I'm interested in seeing the progress.

Originally Posted by Bandit50

Also, have I been talking about this forever? Pretty sure I said last year the project will get started in spring, do you not have patients? If i wanted to give out lectures, I would have made a youtube video and linked it. SO yeah this is public discussion, but you know have some respect before you jump in and hijack it.

This was about drafting, not making a gear box from a ingot of steel, and using a CNC machine. >.> Pretty sure I have said that like 8 times now. Experimenting in CAD, is making something in a simulated 3 dimensional space. It isn't drawing something walking over to a machine uploading it through a USB and test firing after it's assembled what do you think this is 3d print a gun? no what i am going for here.

You're going off on a tangent here. Who is not having patience? Nobody is demanding you do this right now. Only technical opinions have been provided. You are free to ignore technical engineering and manufacturing related opinions as you so please.

As you are surely aware, getting the precise coordinates to the critical features of the gearbox will greatly help in the CADing and reverse engineering process you plan on doing. As you are surely aware, a CMM or a CNC machine with a touch probe can do this fairly quickly and easily, providing that critical info needed to run into a CAD and drawing it yourself. As you are well aware, CMM is often used in conjunction with manual CADing, and its not as simple as plug and play, and replicate.

If you are happy with using a ruler or a hand held micrometer to get your measurements of the critical features of the gearbox to help in your CAD, then more power to you. If you think you need to 'weld aluminum' and 'temper aluminum' for the V2 gearbox clone (versus say machining it out of a single block of 6061-T6 for example), then more power to you.

[Bandit50]

Going off tangent?

Originally Posted by zzzzsleepy8

Again, I could have the V2 gearbox completely CMM and CADed and maybe even machined in max 1-2 days myself if I cared enough to (which I completely don't), instead of talking about doing it for years on end without doing it.

Seems like your impatient about something? What with the
"Talking about it for years" statement.

Yeah, I am aware. Did you even read what everyone else posted before you jump on the alloy topic? It really seems like you didn't.

I have methods, they may not be exactly lab standards, but they will suffice and considering I know what the lab equipment is. Might give you a good idea as to how I am going to do this. The fun of this whole thing, is measuring and drafting it myself. Was that not communicated well or something? I mean this is the 9th time I am saying this now, and the title kind means something to the project. I mean, I know i didn't name this V2 and V3 custom project build.

Why make a machine do it? Where is the fun in that? At least Cactus knew that -.- Great, welcome the level, I guess. ( Also known as understanding where the other person stands on the subject or topic.) Kind of explained within the first 5 reply of this thread.

[zzzzsleepy8]

Originally Posted by Bandit50

Going off tangent?



Seems like your impatient about something? What with the
"Talking about it for years" statement.

Yeah, I am aware. Did you even read what everyone else posted before you jump on the alloy topic? It really seems like you didn't.

I have methods, they may not be exactly lab standards, but they will suffice and considering I know what the lab equipment is. Might give you a good idea as to how I am going to do this. The fun of this whole thing, is measuring and drafting it myself. Was that not communicated well or something? I mean this is the 9th time I am saying this now, and the title kind means something to the project. I mean, I know i didn't name this V2 and V3 custom project build.

Why make a machine do it? Where is the fun in that? At least Cactus knew that -.- Great, welcome the level, I guess.

Lets not get carried away. Again, I think its a neat project. Just providing input. It all started with a disagreement with something you said earlier, about 'being able to easily temper metals as long as its not alloy" along those lines.

The whole point of CAD is precision. It becomes a pointless exercise to hand wavingly measure and CAD gearset locations. Doing so, you are practicing techniques that are no better than those 'crap imports' you were just blasting earlier.

I'm sure since V2 gearboxes were originally designed in Japan, its probably metric so you might be able to round off to the nearest denominator that makes sense while using only hand tools. In any case, you'd best use a CMM to get those coordinates anyway to verify before you CAD it, as I'm sure you know. And as I'm sure you know, CMM isn't just 'getting a machine to do it'. A CMM can grab the entire geometry, but intelligently using a CMM and skilled CADing speeds up the reverse engineering process. Basically you just want dimensions and coordinates of critical features and go straight to CAD.

Anyway good luck.

[Bandit50]

Thank you, you will see some of the basics involved to get measurements and locations of various components. The math involved to find these locations however, will not. That is a well kept secret in the industry that does come from practice and doing. Not from showing

To add, basic principles are the foundations to much of the point, click and navigate devices available in modern society. Measurement is only powerful in the hands of whom can truly use it.

[ThunderCactus]

I know bearings very well, I fix CNC machines so I really have to know a lot about them

I'll let you in on a trade secret, most of the really high machines use 'sealed' greased bearings for ballscrew bearings, tailstock bearings, live tooling bearings, horizontal and vertical spindle bearings.
They are open bearings in packs, but mechanically sealed by cover plates, etc.
Anyway, they last about a decade depending on what you're doing, and they're not "packed" per say, we only load them up to about 15% of their internal volume with grease. So you don't need that much. And that number becomes VERY critical in high rpm bearings. There's been many cases where old millwrights have packed high accuracy bearings 100% full and had them fail after a month. It's because you can't pack them 100% and you absolutely need the proper grease.
Most common reason for bearings to fail are; mechanical impact (that's the big bearing killer), wash-out of the grease (due to seals failing and coolant running the grease out), angular run-out (which is like mechanical wear, running a load in a roller bearing on an angle it's not meant to handle), and overheating (which causes the grease to lose it's lubricity and also decreases the tolerance in the bearings.)

That being said, the high end machines that use high rpm spindles, like 12,000-30,000rpm, use air/oil sprays to lubricate the bearings constantly, but because they are subject to heavier loads for their size compared to airsoft bearings. They need better lubrication.

And airsoft related, what kind of bearings do you think are in the 20,000-30,000rpm motors we're running? They're all sealed, and the very few of them that are actually high end bearings (systema, eagle force, tienly) last a very long time.
With PTW's motors it's the same story, all sealed bearings. Most common bearing to fail in the PTW motors is the top end bearings, most commonly when the bevel is too high, causing heavy loading on one side of the bearing.
The open bearings that most manufacturers use, like G&P, just throw the grease out over time, and if you're dumb enough to use oil, all the lubrication is gone after the first game.

So ceramic sealed bearings will work great for the bevel and spur, but with the sector gear one would think bushings would be best. My concern is the repeated impact with the piston would cause a wear point in the bearing, though I have little evidence to support that. Oddly enough it's usually the spur gear bearings that fail...

[Bandit50]

Originally Posted by ThunderCactus

I know bearings very well, I fix CNC machines so I really have to know a lot about them

I'll let you in on a trade secret, most of the really high machines use 'sealed' greased bearings for ballscrew bearings, tailstock bearings, live tooling bearings, horizontal and vertical spindle bearings.
They are open bearings in packs, but mechanically sealed by cover plates, etc.
Anyway, they last about a decade depending on what you're doing, and they're not "packed" per say, we only load them up to about 15% of their internal volume with grease. So you don't need that much. And that number becomes VERY critical in high rpm bearings. There's been many cases where old millwrights have packed high accuracy bearings 100% full and had them fail after a month. It's because you can't pack them 100% and you absolutely need the proper grease.
Most common reason for bearings to fail are; mechanical impact (that's the big bearing killer), wash-out of the grease (due to seals failing and coolant running the grease out), angular run-out (which is like mechanical wear, running a load in a roller bearing on an angle it's not meant to handle), and overheating (which causes the grease to lose it's lubricity and also decreases the tolerance in the bearings.)

That being said, the high end machines that use high rpm spindles, like 12,000-30,000rpm, use air/oil sprays to lubricate the bearings constantly, but because they are subject to heavier loads for their size compared to airsoft bearings. They need better lubrication.

And airsoft related, what kind of bearings do you think are in the 20,000-30,000rpm motors we're running? They're all sealed, and the very few of them that are actually high end bearings (systema, eagle force, tienly) last a very long time.
With PTW's motors it's the same story, all sealed bearings. Most common bearing to fail in the PTW motors is the top end bearings, most commonly when the bevel is too high, causing heavy loading on one side of the bearing.
The open bearings that most manufacturers use, like G&P, just throw the grease out over time, and if you're dumb enough to use oil, all the lubrication is gone after the first game.

So ceramic sealed bearings will work great for the bevel and spur, but with the sector gear one would think bushings would be best. My concern is the repeated impact with the piston would cause a wear point in the bearing, though I have little evidence to support that. Oddly enough it's usually the spur gear bearings that fail...

It's the impact of the teeth on the piston ram as it begins to compress the spring, I assume we have all seen the damage from poor alinement, cause by the head of the piston not being extended enough and effects the AoE.

[zzzzsleepy8]

Actually, I think its the end stages of the spring compression that is causing higher stresses on the bearings. When you first begin to compress the spring, the force is low and gets higher as you compress the spring further. Hence also why I think cheap pistons have metal racks at the end of the piston (before its released to shoot air out of nozzle) instead of having metal on the full rack.

Sure there is an impulse force upon engaging the rack during the initial spring compression, but the gears probably don't spin fast enough to create as large a force compared to the fully compressed spring pushing back against the gears.

[ThunderCactus]

all pistons have at least one metal tooth on the end of the rack. not because of higher spring tension, but because of the final gear of the sector gear slipping off that tooth. Think of all that built up pressure on the very tip of the tooth, it would very quickly wear it away if it were plastic.

The sector gear with the piston all the way back isn't putting as much force on the bearing as you might think, and not in the spot you might think. Don't forget the spur gear is also pushing against the sector gear.

And as much as we know about what bearings fail in what place, I hate to say it but now that I think about it, none of that actually matters. The bearings that fail are always OEM or cheap replacement bearings.
When a bearing fails in a cheaper end CNC machine, it's not necessarily because something went wrong, sometimes it's just a really shitty bearing.

[pestobanana]

I haven't read the whole thread so forgive me if I'm making a useless post.

Every time I've had a bearing fail, it has been on the bottom side of the spur gear, and never on the sector or bevel. My guess is that is because on the sector gear and spur gear, mechanical contact is made on both ends of the gear, whereas the spur gear only makes mechanical contact on one side. That put's almost the entire load of the spur gear on one bearing bushing.

[zzzzsleepy8]

Originally Posted by ThunderCactus

all pistons have at least one metal tooth on the end of the rack. not because of higher spring tension, but because of the final gear of the sector gear slipping off that tooth. Think of all that built up pressure on the very tip of the tooth, it would very quickly wear it away if it were plastic.

The sector gear with the piston all the way back isn't putting as much force on the bearing as you might think, and not in the spot you might think. Don't forget the spur gear is also pushing against the sector gear.

And as much as we know about what bearings fail in what place, I hate to say it but now that I think about it, none of that actually matters. The bearings that fail are always OEM or cheap replacement bearings.
When a bearing fails in a cheaper end CNC machine, it's not necessarily because something went wrong, sometimes it's just a really shitty bearing.

Pistons that come with AEGs often only have 1 or two rack teeth being metal at the end due to the higher sheer stresses at the end of the compression, again as a result of the higher forces.

Reliable surface-to-surface sliding motion on the last teeth has more to do with hardness mis-match and resilience of both the gear and rack material. So its not necessary per se that the last teeth be metal to provide this capability. Metal is convenient and as that is also the gear of highest forces and since the gears are metal, making the last teeth metal ensures hardness matching. During the slip stage, you go from few tooth engagement during earlier stages (higher contact ratio) to single tooth at the end and with forces applied right at the tooth's cantilevered edge too, so the stresses build up immensely on that single tooth right before piston is released.

Some of the pistons you can buy as an upgrade either have full metal racks, or have partial racks that are more than just the final one or two teeth. The theory I guess for the partial metal ones must be again due to the higher forces as the spring is more compressed at the end of the cycle. So to save weight and cost, they only put metal where it counts.

For support bearings, a lot of times its axial misalignment that leads to premature failure. When you repair a CNC machine's bearings for their ballscrews and what not, how do you ensure they are axially aligned perfectly as well as to their linear guides? Do you put the bearings on first on both ends to allow natural alignment, and then tighten both gradually? Are there measurements that can be taken to verify good alignment? Maybe based on vibrations of the ballscrew?

[ThunderCactus]

Linear guideways are only used in light duty or very cheap machines
Big boy machines for roughing and high accuracy work use box ways. They're more accurate under load and way more rigid, but your maximum rapid travel rate is lower. They also cost a lot more.

The machines I usually work on typically have 3 bearing in each block of the ballscrew, which consists of 3 angular roller bearings stacked like so: <>> ballscrew <<>
First, most important reason, is to absorb the force of stretching the ballscrew. Roller bearings would explode since they're not meant to take the axial loading.
Second reason, is to take up any angular misalignment there might be from, say, some idiot deciding he wants a 30ft bed lathe installed over two separate concrete pads, and over 3 months the floor settles and the lathe ends up having a twist in it's casting.

Now, if you've ever read anything about angular roller bearings, you'd know they need to be squeezed in order to functional accurately. The "pillow blocks" on our machines are machined, and the outer cap sometimes has shims on it to squeeze those bearings to the exact size we need them to be. The reason we've got 3 angular bearings instead of just two, is not just for added stability, but also to help properly press those bearings, the slack ends up between two of the outer races, so when you put the cap on the pillow blocks, the outer races squeeze everything together.

Once the bearings are in and the caps or collars that squeeze the outer races are on, the threaded locking rings go on. They thread onto the ballscrew itself, and push against the inner race of the bearings, that force goes to the outer race, which is pushed against the pillow block.
As tightness is concerned, we put a test dial on each end of the ballscrew and tighten until they stop creeping in either direction. Then there's a formula we use to determine how much to stretch the ballscrew, this is for thermal compensation, the ballscrew heats up during use and stretches out to being normally tight, otherwise you get some wobble. Once it's tightened down, we straighten the locking nut on the ballscrew to remove angular runout. Because the threads on the locking nuts aren't crazy high tolerance, the nut can shift to one side or another by a few thousandths, that translates to angular force on the ballscrew.

That's about all you can do. The ballscrew nut also removes a lot of vibration on it's own.
If you're testing a servo controlled axis for vibration, most the vibes you'll pick up will likely be from the servo, even during normal operation.

[zzzzsleepy8]

Originally Posted by ThunderCactus

Linear guideways are only used in light duty or very cheap machines
Big boy machines for roughing and high accuracy work use box ways. They're more accurate under load and way more rigid, but your maximum rapid travel rate is lower. They also cost a lot more.

The machines I usually work on typically have 3 bearing in each block of the ballscrew, which consists of 3 angular roller bearings stacked like so: <>> ballscrew <<>
First, most important reason, is to absorb the force of stretching the ballscrew. Roller bearings would explode since they're not meant to take the axial loading.
Second reason, is to take up any angular misalignment there might be from, say, some idiot deciding he wants a 30ft bed lathe installed over two separate concrete pads, and over 3 months the floor settles and the lathe ends up having a twist in it's casting.

Now, if you've ever read anything about angular roller bearings, you'd know they need to be squeezed in order to functional accurately. The "pillow blocks" on our machines are machined, and the outer cap sometimes has shims on it to squeeze those bearings to the exact size we need them to be. The reason we've got 3 angular bearings instead of just two, is not just for added stability, but also to help properly press those bearings, the slack ends up between two of the outer races, so when you put the cap on the pillow blocks, the outer races squeeze everything together.

Once the bearings are in and the caps or collars that squeeze the outer races are on, the threaded locking rings go on. They thread onto the ballscrew itself, and push against the inner race of the bearings, that force goes to the outer race, which is pushed against the pillow block.
As tightness is concerned, we put a test dial on each end of the ballscrew and tighten until they stop creeping in either direction. Then there's a formula we use to determine how much to stretch the ballscrew, this is for thermal compensation, the ballscrew heats up during use and stretches out to being normally tight, otherwise you get some wobble. Once it's tightened down, we straighten the locking nut on the ballscrew to remove angular runout. Because the threads on the locking nuts aren't crazy high tolerance, the nut can shift to one side or another by a few thousandths, that translates to angular force on the ballscrew.

That's about all you can do. The ballscrew nut also removes a lot of vibration on it's own.
If you're testing a servo controlled axis for vibration, most the vibes you'll pick up will likely be from the servo, even during normal operation.

Thats just a play on semantics I guess. Box ways and linear rails are linear guides. Smaller linear rails like those supplied by THK or custom box ways used on a Cincinnati.

Interesting. I've also worked on man-handle-able sized ballscrew setups with fixed-free bearing support setup where angular contact thrust bearings were used on the servo drive end with a thread locking nut and a standard radial support bearing on the other end.

Thats an interesting account, thanks. You must be working on those massive CNC machines.

[ThunderCactus]

If it's got a roller bearing on one end, then they're not stretching the ballscrew, but they're probably allowing it room to stretch. Unless you're accounting for that in your software, it'll result in small positioning errors. Our HAAS lathe would change about .003" over lunch as it cooled down.
The angular thrust bearings on the servo side are probably to help dampen vibration and possibly correct any servo misalignment.
I know some machines just use roller bearings on both ends of the ballscrew, but you don't want to buy those machines for serious work lol

I work on anything that has a fanuc controller

But even the 'smaller' mills need box ways if they're doing moderate work, like the DNM500
I ran a HAAS VF3 and new style VF4 for a few years, I can tell you how important it is to have rigidity in the machine. And it makes a big difference even when you're doing moderate machining of steel. I could flex the head of that HAAS just running a 1" indexable end mill at 2100rpm and 50ipm at .125" depth and 70% cutter immersion.

linear rails are for things like;
-making sure large overhead panels slide straight when there's only one actuator that's off-center
-for mounting controllers to so they can slide back and forth on a long bed lathe
-for doing really fast positioning but low load actions like positioning laser cutter heads, or drilling and tapping small parts

They're great for fast and accurate positioning, but they're just not meant to handle loads.
But they're great on those tiny little chinese machines, wouldn't trust anyone scraping those so the linear guides are an advantage. And you're not likely to be doing anything very heavy on them anyway