GD&T for flat bar "twist?" 1

[whotmewory]

Anybody know of a better way - other than straightness - to specify against excessive "twist" when welding to a long, narrow flat?

I know straightness will govern "bowing along the full length, but does it also govern "twist" along an axis tangent to the long axis?

Chris in NC

 

[ewh]

How about "flatness"?

 

[MadMango]

You'll probably get more in-depth responses in forum1103

But I agree with ewh, use Flatness.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[whotmewory]

ewh & MadMango:

I disagree with that. If you recall that flatness is just two parallel planes within which all elements of the surface itself must lie, cannot even the surface elements of a "helix" twisted piece of bar or plate meet and comply with "Flatness?"

 

[ewh]

If you specify flatness tight enough, there will be no "corkscrew". Any geometric control you place on it will be subject to your objection as stated above.

 

[JStephen]

A helix does not lie in a plane- a plane is flat, not an arbitrary curvature.

 

[BillPSU]

When manufacturing on a long narrow part generally there are four different conditions which can occur. The conditions occur when slitting, shearing, forming and welding.

1. Bow the length of the part in the vertical direction

2. Bow the length of the part in the horizontal direction

3. Bow across the width of the part

4. Twist

Other conditions can occur such as a kink or a radical bend but those I classify as obvious bad part which cannot be used unless the localized problem is fixed.

GD&T will usually call out flatness which doesn't care which direction the deformation is but it falls within the tolerance band. Usually you find out deformation in one direction is not as important as the other direction and GD&T has a tough time trying to give a tolerance. Usually notes are required to attempting to clarify what you actually can accept.

 

[MechNorth]

Straightness will only cover you in one direction, not 3-dimensionally. Instead, use flatness as previously suggested, but add a refinement for unit flatness. For example, top line could be 0.5, while the bottom line could be "0.1 / 10X10". Set the top value to limit the overall flatness of the entire piece, and the second line effectively limits the twist rate. Hope that helps.

 

[Tunalover]

Folks-
Aren't you are all overcomplicating the issue? Use a bilateral profile control around the perimeter of the cross-section (being careful that it is less than the size tolerance). By default the control applies over the entire length of the bar.


Tunalover

 

[MechNorth]

Hi Tunalover,

Good suggestion. You can use a bilateral surface profile with a sufficiently tight tolerance, but that may be over-restricting it. If you can live with a twist, but no radical transitions, a two-level flatness with a refinement for unit flatness will do the trick.

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

 

[Tunalover]

MechNorth-
I think you missed this one. Consider a square bar distorted into a rhombus. The rhombus can have four very flat surfaces but you are still not controlling the form of the SHAPE.


Tunalover

 

[MechNorth]

Hi Tunalover,
I think we may be talking two different applications.
I'm working at generating a single surface which has no sudden twist and the overall twist is controlled; consider this surface mounting against another piece and acting as a primary datum feature if needed (maybe a latch bar or similar where only one face is engaged). If the twisted bar is being straightened in a press, then inspection of only one surface is needed. Opposite and adjacent faces could then be controlled to it as needed. In this situation, controlling all faces with a single profile tolerance (all-around) may be over-restrictive and drive costs up by verifying all faces.
If all faces need to be controlled tighter and the entire bar needs to have minimal twist, a surface profile is a great control. This would control it for bowing longitudinally in both directions as well as twisting.
I've had both situations; the first where pure functionality was the concern, and the second where aesthetics were critical as well.

 

[whotmewory]

I certainly appreciate all the contributions to my question.

I brought this up primarily because I am concerned about receiving WELDED fabrications in which [the] vendors insist the parts are "per drawing," even if they twist 1/16 in 4-inches linear. I want to protect our company's interests by communicating to the vendors as best possible, and from the length of this discussion, I'm beginning to wonder if this is something ANSI and GD&T tolerancing methods have inadvertantly overlooked?

 

[MechNorth]

Whotmewory, the controls will do what you want; you just need to be clear and specific about what the functionality of the part is. Does it mount on one face only? Are two opposite faces critical? Are all 4 sides critical? How much twist can you live with overall, and how quick a twist can you accept? The standard doesn't have a specific control for twist, but the controls discussed above will address the twist issue. If you can provide more detail on the bar re function and interfaces, overall size, max'm tolerable twist, and so on, we should be able to provide a couple of solid control options for you.

 

[Tunalover]

whotmewory-
A word of caution:
you're vendor (like many) may ACT like he knows what the drawing says but may not UNDERSTAND what the drawing says. You may want to ask him for certs and for a 100% dimensional inspection report.


Tunalover