[kimfmx]

Hi,

I have purchased a set of liniear rails. And they seem fine, however when I raise them to vertical state (intended for Z-axis), they seem to lock up when beeing mooved. it seems to be more frequent the slower I move them, moving the trains fast up and down seems to go fine. Somehow the balls will cause the trains to lock up to make irattic movement.

I have inspected the Rails, and they look in mint condition, with no markings or sigsn of ball indentations caused by overload.

I have taken the Trains apart and cleaned them, and there are no apparent wear on the trains or ball damage. Everything wiped down lubed and assmbled, and the problem is there. The trains run tight on the rail, so I do not assume its to do with worn out balls. They tend to run much better when used horisontally and under load.

Im afraid that the stepper motor will kism steps using these rails for the Z axis.

Please help.

Best Regards
Kim Mortensen.

[Haydn]

Did you buy new?
I had the same problem with some thk rails I got on ebay. Seems they had been used horizontally and were fine in that plane, I bought them for z and when orientated in that plan they were rough and sticky. I was advised it was because of wear and the change in load direction on the bearings amplified this.

[kimfmx]

No they were also off ebay.

I have looked them through, and they dont seem worn, there are no markings or wobble, and they run tight on the rail.

And I have had everything apart for cleaning.. no difference.. The only think I can think of is the plastic ends (that returns the balls), that could cause the clogging if badly worn.. I have tried to contact the manufacture, to se the replacement end can be purchased for a resonably amount of money.

[LeeWay]

When these things wear just a little, it is hard to see it on really small balls. They get kinda egg shaped. You don't see any wear on the rails and there may likely be none. The rails are harder than the steel balls I think. Some of the football shaped balls are likely being turned sideways while in the vertical orientation and this is what you are feeling. They all lay as they should when on the horizontal and while worn, still appear smooth. You have two choices. You can order new steel balls from many places or you can run them horizontal for awhile and they will have a small amount of play. If you try running will footballs in the vertical, you are going to rapidly increase the damage and may even crush some balls or scare the rails or truck channels.

It's really not hard to put balls in these things. A little tedious, but thats about it. When you get done, make sure you lube everything well.

I have a set 15mm THK's I bought off Ebay a few years back. They were in the same shape. That is what I concluded with those. I haven't yet tried to locate any balls for these, but they shouldn't be very hard to find.

[kimfmx]

Thanks,

Ill try to figure out what size balls are needed, I have already ordered replacement balls for my ball screws - there is a person on Ebay that has a whole lot of them in many different sizes, and thy are sold in bags of 100, and is a fair prize... I guess I need to figure out what size and the amount for each train.

Best Regards
Kim Mortensen.

[kimfmx]

Hmm I have been searching high and low for the ball dimentions of a THK HSR15 wagon, but Its apparently impossible... And I se no use in trying to get balls of the size I have measured from the old egg shaped balls.

From what I could measure, the dimention is about 2.9xx mm But I donet want to ask for 3mm balls if that way to big..

Does anyone know where to find the original ball size of the THK HSR15 wagons. FYI there are 124 balls in each wagon.

Best Regards
Kim Mortesnen

[HuFlungDung]

It might be egg shaped balls, or it might be that all balls are the same size.
Have you ever ridden up behind your buddy on his bicycle, and let your front tire touch his back tire? What happens?

Same thing happens in a ball circuit if all balls are the same size and adjacent pairs are making contact with both the raceway, and the ball in front. This is why spacer balls are used in some ballscrews. A small ball will not be in firm contact with the raceway, and this permits it to run backwards if it accidentally touches the big ball. Because the small ball is small, it clears the raceway and does not skid and cause a lockup.

You may have seen cutaways of some ball trucks which have a plastic cage within, that is flexible enough to move with the balls and also keeps them from touching one another. You might be able to simulate this with short pieces of plastic tubing between, but that would be a total experiment, and could be frustrating if they catch somewhere

[Haydn]

Egg shaped balls !

Yup thats what I had

[fatal-exception]

Kimfmx, you might be able to get a more accurate measurement of the ball size if you lay 20 or more in line then mesaure the total and divide by the number of balls. I did this with one screw that I had, ordered the 'theoretical' size, and the refurbished ball screw has been in production for over a year, still no backlash.

Haydn, could you tell if the balls were sliding over the rails rather than rolling?

[LeeWay]

The football shaped balls I have are easier to feel rather than see. I checked them with some 6" digital calipers and there was a definite size difference depending on where I measured them. I don't recall how much it was though.
I think you might get better measurements using a micrometer. They will lay on the table like a football would with the long ends pointing outward. Just keep measuring until you get the big sides. Measure several of them also. Use the largest measurement. I think you can set these trucks up like a ball screw for preloading. Use some smaller balls and some large balls alternating each one. The size you measured should be good. Perhaps just a hair larger to preload them. I would try to get as close as possible to that size though. Then for the smaller size, get close to the same size but under what you measured.

I measured the preloaded ball screw balls from my ground NSK ball screws and they were 3.14 mm and 3.09 mm. Not much difference there. These screws have no detectable backlash and they now interpolate perfect circles. I like that.

I think you will have to trash all the balls in the trucks. I doubt any will be reusable.

[Haydn]

Kimfmx, you might be able to get a more accurate measurement of the ball size if you lay 20 or more in line then mesaure the total and divide by the number of balls. I did this with one screw that I had, ordered the 'theoretical' size, and the refurbished ball screw has been in production for over a year, still no backlash.

Haydn, could you tell if the balls were sliding over the rails rather than rolling?

I think the problem with mine was they would run fine horizontally with a load on, when orientated vertically for z the carriages would drop smoothly then get stuck, when applying a load in this plane they still got stuck but would free with a little force. I suspect that the balls were out of shape, would run smoothly then rotate on the return track bend (if you get my drift) and get stuck. Hard to say if they were sliding or rotating, but i'd imagine rotating or they would feel very rough.

[CNCAddict]

Egg shaped balls? I have never heard of that before. If they are "egg" shaped then the block would run terribly in any position. In the vertical orientation the balls must be PUSHED up a very long vertical pathway. If you have long blocks this could be a very big problem especially if there is a lot of wear in the plastic return section of the block. This would allow the balls to stack in a zigzag pattern and cause some pretty bad jamming. If you can't measure play in the rail, then getting larger balls will just cause an increase in preload of the rail if you can even get the block onto the rail. Too high a preload and things will run very rough and decrease the life of the rail. Either way it won't help the problem you have. I'd try to inspect the return path to see what it looks like...maybe you could shave off a little material to bring it in closer and tighten up the clearance a bit??

[schbrownie]

Couple of things... nobody has mentioned the obvious yet... the first thing that could be wrong is misalignment/binding in the Z-axis. Have you double checked to make sure that your system is perfectly alligned?

Secondly... almost every profle rail manufacturer in the world has different preloads of carriages (aka sliders, trucks, etc). In order to obtain the proper clearance or interference fit, they manufacture the rail to the best tolerances they can hold (usually +/- 0.0002"). They then grind the carriages and because you're grinding an internal feature (which is harder to hold tolerance), they're probably only able to hold +/- 0.0005". So, if you have a tolerance stackup of +/- 0.0007" and you need to control this stackup to be less than 0.0001" as a system, how do you do this? Easy, each carriage is measured and then the proper ball size is selected and installed into the carriage to ensure proper fit. If you ever go to a trade show and see a machine which makes bearings automatically, you'll see several towers of ball bearings on the machine and each one is marked (-10, -9, -8, ..., -1, 0, +1, ..., +8, +9, +10) and each tower contains balls that are slightly different in size from one another.

From Hiwin's catalog (because I have it in front of me), for the Z-axis, you'll want a heavy preload for high rigidity with vibration and impact. You'll know if you have the correct preload becuase the carriage won't want to move easily on its own. It's possible that the rail you got from eBay was a non-preloaded or lightly preloaded system... which is ideal for horizontal systems.

Someone previously mentioned to use a caliper to check the ball diameter... this will not be your best option. Balls used in these systems (even by the lowest cost manufacturers) only vary by 0.0001" from one size to the next. A caliper is only accurate to +/- 0.001" (usually), so there's no way you can accurately measure to one tenth. In order to do this, you'll need to use a micrometer. Also, while stacking 10 balls in a row and measuring them is an ideal way to average the size, you still run into the same problem if you use a caliper...

[guru_florida]

Further to schbrownie, what model of rails/trucks do you have? As he was saying, some are intended for horizontal only. The forces on z-axis, specifically moment forces, are very different than horizontal.

C

[kimfmx]

I have not recieved the notifications for all those answers, but thanks for the answers...
The Trains are THK HSR15. I have now dfound some others on Ebay, and is trying to sell these off here locally (Denmark) for horisontal movement only.

I am also more into the wear on the end return path, as this is plastic, and would wear faster, and I could very much se the zig-zag action when placed vertically. I dont intend to repair these, as I could only imagine what the spares would cost...

I tried to order new balls for my leadscrew, that had some little sideway movement, but I can not seem to measure any backlash... Is it possible that the leadscrew is work in the depth, and not on the sides of the bearing grooves - From a design perspective, the lead screws were only made for presition along the axis, and not perpendicular to ....

[vger]

Unusual thought about how to determin what the original size of the "egg shaped" balls are. Take 50 or 100 of the balls, and measure the volume of them all together. Take graduated cylinder with light oil to a defined level, add the balls, read the new level, divide the change by the number of balls. Then just calulate the diameter by the volume.

radius = (volume * 3/4pi)^1/3

Might work

Steve

[LeeWay]

It would depend on how the balls got contorted to begin with. If they were simply mashed and compressed, then it might work. If they were ground any because of loss of proper movement, then there will be lots of smaller particles that ain't gonna make it into the measurement.
In my case, I think it started by compression and was compounded by wear.
I bought some steel balls from Mcmaster Carr. $2 something per hundred count. Just standard 1/8" balls, but good to have around for testing and such.
The high precision balls of high quality screws and slides are likely not standard sizes though.
Similar to a lock keying set, to do this correctly, you would need a set of multiple ball sizes all in nearly the same range with variances of sizes in the mils.
I measured my precision screws (with digital calipers) @ 3.09 mm and 3.14 mm. Even with calipers, I could tell there was a difference in sizes. A mic would get you closer to perfection of course, but I am not sure we need to get that precise especially with linear bearings and rolled ball screws. Perhaps with preloaded ground ball screws, it would be prudent to use a mic and get the exact sizes needed.