alignment table/build surface tolerances

[prolix21]

I've been shopping around for a decent surface plate or alignment table for a surface jig I have. What sort of tolerances are generally necessary/acceptable on a table for frame building? I have a friend who has a fab shop and we've been talking about building something interesting ourselves but we're not sure of the tolerances we need to be within. The various surface plates I've looked at have pretty incredible tolerances, but then I hear stories about alignment tables at various factories having tolerances with +/- multiple millimeters. That's far from the thousandths that most surface plates seem to be (granite or otherwise).

I recall e-richie saying something about his measurements being only reproducible on his alignment table, which was also the story at UBI (same model of table I believe) and in a lot of cases you couldn't even duplicate them consistently if you sat there all day flipping the frame around. When all was said on my first frame I was somewhere within 1-1.5mm of runout when I put it on the table. That was averaged out over a few times of taking it on and off the table and that's assuming the table was accurate. After fitting up the bike it seems a ton straighter than anything else I've ever owned.

Anyway, just curious what the opinion here is. And yes, buying a chunk of granite would be easier, but where's the fun in that? :)

[sam]

I dont have a direct answer for you, but a general answer would be as flat as possible, if my standoff/bb tower, etc. say is built .005 off, and the table where it sits is .005 off, i might now be .01 off, and .01 across a 56cm tube can add up to a fairly big number. so not only does the table have to be flat but everything in the process does, as the numbers stack up. I see alot of shop made tables that look nice but in time are going to warp, so if you dont have the experience in building a precision table you are better off buying one.

That being said, I have also seen some tables from some respectable builders that I can see are also not the best tables,but they know how to work with what they got so it dont matter as long as the frame is straight, so in some thoughts people play to much into how flat there tools are. But if you are going to do it do it right, but I think you are going to have a hard time getting someone to say here this is the number.

[Andrew R Stewart]

Sam touches on one of the issues that i've had with discussions about frame alignment. People are quick to throw out numbers but few place said numbers in a frame of reference that is needed to make full sense what the number means. So a .005" offness is what? Is this over a few inches of the Bb post base or is this along 10" of seat tube or across 20" of a head tube alignmeng bar? Major differences here in real life applications.

Another aspect that is less often talked about is the precision of the tubes themselves. Ever roll out or measure diameter over the length of a tube? Then after brazing you can see distortions in tube diameters close to the joint. So where do you start the measurement? If you're using a BB face as the starting point (as on a whipping post) what about it's distortion (that much has been said about). Just because with a dial indicator you can see .001" difference is no reason to think it is an important amount.

I try to get a picture to the precision of the tubes first and realize that I'm not going to get much better then this limit. I measure alignment a few times on each side of the bike to further get a feel for the frame's center line alignment. Currently i hold the frame by the shell on a post (but soon to go to a head tube held method) and measure the HT to the plate and then the ST to plate over the same length. This way i seperate out the shell's alignment from the frame's tracking alignment. Over the same length of HT/ST i try for .015" of difference or less, this length is usually 9"-12".

Maybe as important to any ultimate alignment is the ability to be consistant in how you measure and then to be able to repeate to a degree the measurement. When I had my own shop I would do a number of Winter overhauls, most of these bikes got a measuring session on my plate. Not so much to sell another service, backed up by numbers (and who can argue with a number?) but more to add to my personal data bank and learn how setting up the measurement can effect the read out. You'd be impressed how off so many frames are and the rider never knows it.

So some of your answer (to the not directly asked question) is that you have to build an understanding as to what's enough, alignment wise. Your surface plate needs to be at least as good as this personal tollerance limit and it has to be stable enough to be able to repeate the measurement (with in some reason). If your plate is off by more then your personal tollerance limit then you need to know that and know how to accomidate that.

So there's no one answer that we all will agree with. But then you knew that. Andy.

[prolix21]

So there's no one answer that we all will agree with. But then you knew that. Andy.

Yeah, I think I knew that already too :)

I heard a story when I was out in Oregon about one of the alignment tables at a rather big bike company reading that the table itself was something like 5mm off so who knows how off the frames ended up being.

Something I've had emphasized to me by other frame builders is that these are 'hand' built frames and there's only so much you can do to overcome the nature of the material and the process (mainly, heat). I guess I need to figure out what that personal tolerance is and build accordingly.

I should also stop over-thinking this whole thing and start another frame.

[e-RICHIE]

I should also stop over-thinking this whole thing and start another frame.

It may seem counter-intuitive and counterproductive, but this ^^ is exactly what you should do atmo. The very task of building a frame - learning to attach pipes to other pipes - as well as the design parameters you work with (fit, handling, clearances for components...), should be your focus. If you don't know, don't learn, or can't do these first, all the precision in the world is worthless. There are a lot of really nice (and heavy, and large...) systems being shown on several boards and I think the fellow(s) with the torch and files is/are putting carts before horses.

[ron l edmiston]

It may seem counter-intuitive and counterproductive, but this ^^ is exactly what you should do atmo. The very task of building a frame - learning to attach pipes to other pipes - as well as the design parameters you work with (fit, handling, clearances for components...), should be your focus. If you don't know, don't learn, or can't do these first, all the precision in the world is worthless. There are a lot of really nice (and heavy, and large...) systems being shown on several boards and I think the fellow(s) with the torch and files is/are putting carts before horses.

so true mr sachs...

in motor sports ...,
building a road racing vehicle means understanding chassis and suspension --- if brute horsepower is your main stroke, the drag strip maybe your calling card..,
your horse "without the cart.."

ronnie

[prolix21]

It may seem counter-intuitive and counterproductive, but this ^^ is exactly what you should do atmo. The very task of building a frame - learning to attach pipes to other pipes - as well as the design parameters you work with (fit, handling, clearances for components...), should be your focus. If you don't know, don't learn, or can't do these first, all the precision in the world is worthless. There are a lot of really nice (and heavy, and large...) systems being shown on several boards and I think the fellow(s) with the torch and files is/are putting carts before horses.

In some ways I think my first frame experience put the cart before the horse. I certainly learned a lot at UBI and I walked away with what I consider a successful frame. It's solid, straight and everything fits fine, but I wish I had struggled more and in some ways I wish I would have fallen on my face a few more times than I did. Perhaps more to the point, I wish I hadn't learned with such great tools and such accurate fixtures. I feel like I skipped ahead and lost something in the process and now I'm sort of reverse engineering things and trying to build frame number two by forgetting all the tools and crutches I had on frame 1. I'm probably learning more through this process, and will no doubt learn a ton more by the time I complete frame 2, but its hard not to think that an anvil jig would cure all my woes.

Part of the reason I'm going with a flat plate fixture is to get a more basic/manual process and a feel for how it all fits together.

[Jacques]

Putting asside the question of frame building best practice (I'm a machinist, not a frame builder) flatness tollerance is one of those things where if you have to ask, you probably won't know what to do with the answer either. I'm not saying this to belittle anyone, just saying that the information has little utility to anyone without access to a metrology lab.

Flatness is the measured deviation of a surface from a perfect plane. A flatness tollerance is the distance between two parallel planes encompassing every point on the surface. In order to inspect for flatness, you need a plane (approximate, of course) that's as big as or bigger than the surface being inspected.

All of this being said, knowing the flatness tollerance of a surface plate still tells you very little about the possible margin of error when measuring on it because it says nothing about the distribution of those high and low points. Add to that the variability in shape and size of the instruments and parts that come into contact with it and you see why it's hard to say what degree of precision is required.

And in case this post wasn't anal enough, runout is used to mean only the deviation of a circular feature from a part's reference axis.

[Eric Estlund]

We are well into thread drift now, so I'll chime in.

Tools are only as helpful as their operator's process and implementation allows. This is true for the tooling at ubi,Doug's, or my shop as it is for your granite/iron/steel table. You have been introduced to a method to produce a bike. The next step is to mentally build a bike and map out the process and operations (not the tool stations). One you have a visual path, assess a realistic physical process.

If you need to slot stays, the bike doesn't care if you do this with a mill or a hack saw, so long as ir gets done. Be careful with conflating the operation (slotting) with the tool (mill). Same with workholding (jig, vise, gravity), etc. What needs to happen doesn't change much, and the "best" how will evolve with you.

To bring this back to the plate- you need to get comfortable with what/ why you are measuring, and what if anything youwill do with that info. Once you have answers for how the measurements are relaventfor you, the specifics of the tool will be more obvious.

I hope that helps- one can make crap bikes with a space age shop, or fantastic bikes with little more than a file. Don't buy more than the file untill you know what it will do fir you.

Xo-from a shop that is still virtually all hand tools (and a heavy piece of flatish steel).

Ps- practice, practice, practice.