Proper way to address flatness

[gsepan]

Good Morning all,

My question is regarding a flatness callout on a print I'm working to. The callout wants to hold a flatness of 1.5mm on a mating surface of a sheetmetal part. I took a series of 12 surface points on the mating surface, created a plane from these points and then dimensioned flatness. Is this the proper method or am I doing something wrong here? How do you folks accomplish this task?

Best Regards,


Greg

[bw_bob]

Flatness is to be checked in the free state, not restrained.
Then take the # of points which you are ok with to create the plane.
1.5 flatness if inches is a pretty wide zone.
Dimension flatness & your done.

[Thomas Stewart]

Yeah gsepan, you're right on track. Flatness is one of the tolerances that are relatively simple to create. Although 12 points may be a bit accessive, it depends on the size of the area and potential NON flatness you're dealing with.

Although Flatness can be A bit decieving in that what we're doing here is creating a "best-fit" scenario and truly is one point is out on flatness while the rest are in (in varying degrees), it will generally show IN-tolerance.

I typically like to evaluate each point independant before creating a plane. I won't necessarilly tolerance those individual points, but it'll give me an initial idea as to what I'm dealing with or whether the surface is actually flat or simply falling into the "best-fit" surface...

[gsepan]

Thanks bob,

Seems like I'm doing the same thing that you do when it comes to checking flatness, I just wasn't sure of the readings I was getting. Do you know how pcdmis calculates flatness? Does it matter in what order your points are selected when creating the plane? I haven't checked yet but I think that I would have much worse results if I checked with an indicator on a surface plate.

[bw_bob]

Thanks bob,

Seems like I'm doing the same thing that you do when it comes to checking flatness, I just wasn't sure of the readings I was getting. Do you know how pcdmis calculates flatness? Does it matter in what order your points are selected when creating the plane? I haven't checked yet but I think that I would have much worse results if I checked with an indicator on a surface plate.

I believe flatness is calculated by the min/max of your points.
The order in which you take them does not matter.
You should get worse when checking with indicator. Your checking the whole surface. The CMM is checking only the points you've took with it.

[gsepan]

OK now I'm a little worried... I asked for the profile of a surface on all 12 points and I looked at the range of the readings and here is what I found. One of my points show a deviation of -0.57mm and another shows 0.52mm. So the range inbetween the two is 1.09mm. Pcdmis tells me the flatness is 0.56mm...

[Matthew D. Hoedeman]

Flatness does NOT USE POSITIONAL information for it's calculation. A plane is a plane is a plane and can be ANYWHERE, as long as it is flat, that is what flatness is for. Using PROFILE OF A SURFACE will include the absolute position of the points.

The plane can be off-angle by 30 degrees and still be perfectly flat and your flatness will tell you that it is flat and in tolerance. Flatness is NOT LOCATION and should not be confused with location (profile).

[Roger Kilpatrick]

--------------------------------------------------------------------------------

I will NEVER check flatness with a CMM, I bought a 2" thick surface plate, had a hole cut through it and fabricated a holder for a digital gage sender unit that sticks up about .05 through the plate. I lay the part down on the plate and run it over the indicator sending unit tip checking flatness of the surface while the part rests on its high points. We never had a way of accuraely checking flatness per ASME Y14.5M prior to that. This method works great.

[gsepan]

Thanks to all...

Roger, sure would like to see a picture of your method of checking flatness. If you could post it for all to see, that would be great!

Matt, thanks for setting me straight, I've got a better grasp now with your insight.

Have a great day everyone!


Greg

[Roger Kilpatrick]

FLATNESS.jpg

This is what we use for flatness, home maid but effective.

The sender is located in the center of the plate in this pic, but if you look to the right you can see another holder the sender can go in, this way we can check flatness on a stepped surface while the rest of the part hangs over the side of the rock.

[djayne]

I have had problems with flatness in the past and present. I have a strut mount with a min surface finish of 25Rz and have tried all kinds of methods of measurements to get repeatable flatness results. It's pretty much out of spec all of the time - If it is ever in spec, we immediately check the surface finish to see if it is too smooth. I never was quite sure if it was my measurement technique or the roughness of the surface. I am planning on looking at some flatness gages at IMTS in a 2 weeks.

[Matthew D. Hoedeman]

Have you tried a BIGGER diameter probe? If the surface needs to be rough, then you need to try to hit the high points to make the plane. I would try a 6mm ball probe, that will help keep you out of the valleys between the mountains, so to speak, and will give you flatter results.

[bboyd]

I've never been a fan of using the CMM for checking flatness or any other form tolerance, but have done so when there was no way to isolate the surface and use an indicator.

[Tested]

--------------------------------------------------------------------------------

I will NEVER check flatness with a CMM, I bought a 2" thick surface plate, had a hole cut through it and fabricated a holder for a digital gage sender unit that sticks up about .05 through the plate. I lay the part down on the plate and run it over the indicator sending unit tip checking flatness of the surface while the part rests on its high points. We never had a way of accuraely checking flatness per ASME Y14.5M prior to that. This method works great.

We have done basicly the same thing but use a air gage rather than a indicator.