I have an aluminum plate that will have a smaller, but thicker aluminum block mounted to the top of it. When the block is mounted to the plate, I need the bottom surface of the plate and the top surface of the block to be flat within .0005. Do I just show the assembly of the two parts with leaders pointing to both surfaces and show the flatness symbol with the tolerance?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Flatness between two parts
- Thread starter Dragon7
- Start date
CheckerHater
Mechanical
Don't mean to appear impolite, but are you sure you are looking for Flatness and not Parallelism?
Unless you're machining them at the assy stage, don't show in an assy, make separate part drawings.
I agree with CheckerHater, seems like you want parallelism. If you want the two mating surfaces to be flush, use flatness.
Chris
SolidWorks 11
ctopher's home
SolidWorks Legion
I agree with CheckerHater, seems like you want parallelism. If you want the two mating surfaces to be flush, use flatness.
Chris
SolidWorks 11
ctopher's home
SolidWorks Legion
- Thread starter
- #5
CheckerHater
Mechanical
In general, you can apply Flatness to several features (surfaces) on your drawings. It will mean you want them to be FLAT, but it will NOT imply any relationship between the surfaces.
If you want one surface to be "flat" in respect to other surface, then what you are looking for is probably Parallelism. (In some cases Profile)
Could you tell us how are you planning to check the condition in question?
If you want one surface to be "flat" in respect to other surface, then what you are looking for is probably Parallelism. (In some cases Profile)
Could you tell us how are you planning to check the condition in question?
Dragon7,
Flatness is a self-contained tolerance that does not reference a datum, on the current part, or any other part. You specify flatness on each of your fabrication drawings. You do not call it up on your assembly unless there is something your assembler can do about it.
Perhaps what you want to do is show the profile or parallelism of the top of your block with respect to the bottom of your plate. This is a valid thing to do on your assembly. The assembler either can control the tolerances of the individual parts, or they can assemble your thing and machine the top of the block.
Let the vendor figure out the interface between the plate and the block. Don't ever take it apart.
JHG
Flatness is a self-contained tolerance that does not reference a datum, on the current part, or any other part. You specify flatness on each of your fabrication drawings. You do not call it up on your assembly unless there is something your assembler can do about it.
Perhaps what you want to do is show the profile or parallelism of the top of your block with respect to the bottom of your plate. This is a valid thing to do on your assembly. The assembler either can control the tolerances of the individual parts, or they can assemble your thing and machine the top of the block.
Let the vendor figure out the interface between the plate and the block. Don't ever take it apart.
- Thread starter
- #8
As I figured, we don't really have a way of testing this until we use it. We're sending the work out and indicating the required tolerance. This is for an RF welding application and we would have to run the machine to see if it was all level.
I had a feeling flatness didn't apply and that's why I came here to ask. I appreciate everyone's input.
Thanks,
Dragon7
I had a feeling flatness didn't apply and that's why I came here to ask. I appreciate everyone's input.
Thanks,
Dragon7
Dragon, your graphic shows us what you have on the drawing, but doesn't tell us what you want by way of functional relationship between the two surfaces.
To clarify what flatness gives you; the top surface could be +65 degrees from the horizontal (as established by the bottom of the plate) but perfectly flat.
What your graphic implies is that you want the two surfaces to be parallel (as suggested several times, above). GD&T parallelism is a little different from high school parallelism, but the core idea is the same. In GD&T, a surface (or axis) has to remain equidistant from a datum (plane or axis).
To start, as indicated above, there are two scenarios to consider; first, that you are somehow machining the top surface after welding. In this case, make the bottom surface of the plate your primary datum feature on the weldment. The top surface can be located by a size with a tolerance (which will automatically give you a certain degree of parallelism), and can be refined with a tighter parallelism control wrt the primary datum. Alternately in this same scenario, you can use a basic dimension to locate the top surface from the datum feature, then use a surface profile to control the location (and to the same extent, the parallelism) wrt the primary datum. You can use a composite profile control with a refinement to give you parallelism, or you can add a separate parallelism control below the profile control to achieve the same thing.
The second scenario is that the two pieces are completely finished before welding. In this case, the tops and bottoms of both pieces will need to be controlled the flatness and parallelism of each surface more tightly than the overall flatness you initially spec'd by distributing it based on manufacturing capabilities. So, for the base piece; the flatness on the bottom surface can be the full initial limit as it doesn't affect the parallelism of the opposite face, and the opposite face parallel to the bottom face within 1/3 of the original flatness. Then, on the second piece, the primary face needs to have a tighter flatness tolerance; again maybe 1/3 of total original, and the opposite face being parallel (and therefore flat) within 1/3 of the original spec, wrt the first face. The distributions are subject to your capabilities, but you get the idea; the total parallelism and flatness errors cannot exceed the original spec.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
To clarify what flatness gives you; the top surface could be +65 degrees from the horizontal (as established by the bottom of the plate) but perfectly flat.
What your graphic implies is that you want the two surfaces to be parallel (as suggested several times, above). GD&T parallelism is a little different from high school parallelism, but the core idea is the same. In GD&T, a surface (or axis) has to remain equidistant from a datum (plane or axis).
To start, as indicated above, there are two scenarios to consider; first, that you are somehow machining the top surface after welding. In this case, make the bottom surface of the plate your primary datum feature on the weldment. The top surface can be located by a size with a tolerance (which will automatically give you a certain degree of parallelism), and can be refined with a tighter parallelism control wrt the primary datum. Alternately in this same scenario, you can use a basic dimension to locate the top surface from the datum feature, then use a surface profile to control the location (and to the same extent, the parallelism) wrt the primary datum. You can use a composite profile control with a refinement to give you parallelism, or you can add a separate parallelism control below the profile control to achieve the same thing.
The second scenario is that the two pieces are completely finished before welding. In this case, the tops and bottoms of both pieces will need to be controlled the flatness and parallelism of each surface more tightly than the overall flatness you initially spec'd by distributing it based on manufacturing capabilities. So, for the base piece; the flatness on the bottom surface can be the full initial limit as it doesn't affect the parallelism of the opposite face, and the opposite face parallel to the bottom face within 1/3 of the original flatness. Then, on the second piece, the primary face needs to have a tighter flatness tolerance; again maybe 1/3 of total original, and the opposite face being parallel (and therefore flat) within 1/3 of the original spec, wrt the first face. The distributions are subject to your capabilities, but you get the idea; the total parallelism and flatness errors cannot exceed the original spec.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
Yes. Unfortunately some companies use weldment drawings as component drawings also. Obviously not a good practice, and they blindly (ignorantly?) pay the price repeatedly.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
Frank,
A weldment drawing is a final product definition. Components are often distorted in some way during welding. So, if the drawing reflects the component in the post-weldment state, then what is the initial state? How does the fabrication shop compensate for the reality of welding/thermal issues in the component?
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
A weldment drawing is a final product definition. Components are often distorted in some way during welding. So, if the drawing reflects the component in the post-weldment state, then what is the initial state? How does the fabrication shop compensate for the reality of welding/thermal issues in the component?
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
Slippery slope there MechNorth.
You could extrapolate that thinking to machined parts - what size billet should they start with etc.
You could extrapolate that thinking to machined parts - what size billet should they start with etc.
Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
What is Engineering anyway: faq1088-1484
If it's a slippery slope, Kenat, then my last question should be easily answered.
I don't see the extrapolation as valid; one is a metal removal process with predictable results whereas weldment is a metal additive process with a greater degree of unpredictability. My thoughts anyway.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
I don't see the extrapolation as valid; one is a metal removal process with predictable results whereas weldment is a metal additive process with a greater degree of unpredictability. My thoughts anyway.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
I don't know, I've seen machined parts warp and twist for one reason or another, that's where I was coming from.
Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
What is Engineering anyway: faq1088-1484
CheckerHater
Mechanical
Ken, Jim,
We all agree that drawing represents the FINAL state of the part, right? So all the twisting and warping should be taken into consideration and controlled using GD&T, isn't that the idea?
It is pretty common (for better or for worse) to combine fab and machining on one print. Then notes like "before weld", "after weld" can be used (and I've seen them being used). Works well, not the end of the world.
We all agree that drawing represents the FINAL state of the part, right? So all the twisting and warping should be taken into consideration and controlled using GD&T, isn't that the idea?
It is pretty common (for better or for worse) to combine fab and machining on one print. Then notes like "before weld", "after weld" can be used (and I've seen them being used). Works well, not the end of the world.
Haven't worked with that many weldment drawings, but none of them had before/after annotations. Would make sense, but it still doesn't help with the warpage allowances, etc., which were big issues and part of the reason I was involved. To me, it seems poor engineering when a weldment has to be ground or milled afterward to meet the raw material (component) flatness/paralellelism spec.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
- Status
Similar threads
- Question
- Locked
- Question
- Locked
- Question