Flatness tolerance machined plate 1

[valmeidan]

I have a plate that I want machined that will have fins on both sides. A lid will then be on either side, and sealed from leaking wit an o-ring around the fins. In order to do this I need a really tight tolerance on the flatness. The part would be about 5mm thick once cnc'd, and the dimensions are 190mm x 540mm. Does this thickness make sense, and what kind of flatness could be achieved. Thanks

 

[MintJulep]

Maybe a sketch would help explain what you are trying to achieve.

 

[valmeidan]

something like this

 

[KENAT]

What material?

The shape will make grinding a bit more difficult.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[valmeidan]

It would be copper, however any suggestions on the type of copper would help also. I know C145 copper is a free machining copper, but not sure if the heat transfer is as good as say CDA110.

 

[IRstuff]

Your base plate is probably going to be too thin to worry about flatness; the covers will warp the base plate into submission. You can lap the flanges on the covers, which would result in an extremely flat and smooth surface. From your picture, the base plate is about 1.5 mm?

I don't get what the purpose is, though. It looks like a heat sink, but covering it up pretty much nullifies any convective cooling.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[valmeidan]

it is a liquid to liquid heat exchanger, and both sides will have a cover with fluid passing through, and no not that thin, i am trying to figure out the thickness required to get a nice flat surface. At 1.5mm thick yes it would warp, but I was thinking around 8mm thick. I was wondering if you could machine that size of surface to have a tight tolerance on flatness say 0.1-0.2mm or less

 

[3DDave]

The better questions to ask:

1) How much warpage can the design tolerate
1a) How much deflection can be generated by the fastening scheme
1b) How much compression variation can the o-ring tolerate
1c) How is variation split between this part and the mating parts

2) How much is this going to cost

You could carve the negative version out of the block of material and put the o-ring groove into the faces of this piece so that they remain as flat as when you paid for them and the mating parts would not have fins to prevent making them flat either.

 

[valmeidan]

i would like to see a 0.1mm flatness tolerance. If there is warpage then the oring may not seal. o-ring compression around .3mm or 25%. The mating parts could be plastic. Is it possible to get a flatness spec around 0.1mm over the given distance of 520mm, if the plate is thick enough, and what thickness would work best to prevent warpage, or is the flatness more a function of the capability of the cnc machine?

 

[MintJulep]

0.1 mm flatness over 540 mm for a big block of copper with extensive material removal on both sides.

Not going to happen.

 

[valmeidan]

is there a spec for this online I could cross reference as far as flatness capabilities based on material and material size? what kind of tolerance would be realistic? The reason for the flatness is to ensure the o-ring around the fins has equal compression and will not leak

 

[racookpe1978]

You don't seem to understand that that thin a plate of copper (a soft metal in the first place) is going to warp across the whole plate dimensions much more than it will leak between the o-ring and the two surfaces o-ring surface. See, between the screws holding the plate down to the other surface, the plate will bend up (and away from the o-ring) far more than that 1/10 mm flatness you are specifying.

Now, machining the copper to that smooth a finish is just barely possible, but you've been told above several times that it will be very, very expensive to hold those tolerances.

Can you use anything more rigid than a thin sheet of copper foil?

 

[MintJulep]

Just go buy a brazed plate heat exchanger of suitable capacity and be much happier.

What you are trying to do will fail without doubt.

 

[IRstuff]

"0.1 mm flatness over 540 mm for a big block of copper with extensive material removal on both sides."

could be done, but it would be VERY expensive, as we once found out when someone did something similar. We had to have the covers lapped, AFTER we had it mostly built.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[valmeidan]

might not be possible based on the server system it will be going into anyhow now that I look at it more closely. Thanks for all your input.

 

[racookpe1978]

Please! Don't worry about failure. It means that you have figured out what does not work.

The real world is an insensitive place. It really doesn't care what your hopes or plans were, but marily goes about its future.

See also:
Panama Canal - French version.
Pyramids - first Egyptian version (stepped pyramids).
Nebuchadnezzar: Rules for killing engineers whose buildings failed, killing of.
Liberty ships - Welding, failure of welding under extremely cold weather.
Hoover dam - Death by carbon monoxide, workers in bypass tunnels.
Transcontinental railroad - Nitroglycerine, explosive nature of when dropped
Hotel dance floor - Vibration and holding rods, changing design position of bolts

 

[dvd]

Non-uniform cooling will create a temperature gradient which results in unequal thermal expansion and warpage. It might be flat some of the time when it is not in use.

 

[tbuelna]

Judging from the drawing of the heat sink and the height of the fins, it looks like any housing attached to each side will have enough depth to make them much stiffer than the heat sink flange. You would want to make the heat sink flange as thin as practical so that it easily conforms to the housing flange surfaces. Also make sure to space your housing flange fasteners close enough together to prevent leakage at the clamped interface due to bending in the flange from internal pressures.

Of course, due to the stiffness provided by the fins, the heat sink body will remain fairly flat in the longitudinal direction within the area covered by the fins.