Simple crush washer 5

[foglght4]

I have to use a copper crush washer to seal against coolant. The bolt is an M12x1.5 simple DIN 908 zinc plated bolt. There is a small interference of 0.1mm between the copper washer and the bolt (worst case), where the bolt surface starts to radius. I need to somehow model or calculate the crush of the material and ensure a proper seal. I'm having even a hard time finding any data supporting the amount of clamp load needed to seal this joint in the first place.

I'm not quite sure I can model the plastic deformation of the copper accurately given pen and paper (excel too I suppose).

Even on just a simple crush washer gasket with a single bolt, I'm not coming up with much in terms of basic equations to understand the clamp load needed to seal the joint using a simple copper washer gasket. How much do I need to compress the copper to fill the voids in the mating materials to ensure a proper seal, or am I thinking about this the wrong way?

 

[MikeHalloran]

You have to stress the copper to yield in direct compression.
That's why copper crush washers are typically modest in their radial dimension.

As for crushing it enough to fill voids of arbitrary size, the calculation is probably a waste of time. In practice, the mating surfaces have to be very close to perfect, and most practitioners will also use a sealant even if none is specified.




Mike Halloran
Pembroke Pines, FL, USA

 

[foglght4]

Right, I know I have to yield it, but by how much to seal?

 

[MikeHalloran]

Well, banjo bolts are normally tightened to 'a little less than what the torque wrench was reading when the first one broke'. Did I mention that I hate banjo bolts and copper or aluminum 'seals'?

I don't know if they still do this, but in the Sixties, Ford used copper washers under the heads of the nuts holding the differential carrier to the 'banjo' of the rear axle, saving them the cost of spotfacing the sand cast carrier face around the bolts. The nut torques were sufficient to hugely distort the copper washers, and also to provide a sufficient pressure on the gasket sandwiched between the carrier and the banjo.

To sort of answer your question, I think the torque to yield the copper is just a minimum. You have to develop enough strain in the bolt (considered as a spring) to maintain that yield force. Then the washer will 'belly out' until the stress falls to just below the yield point.

How about omitting the copper washer and substituting a gob of Loctite Pipe Sealant with Teflon?


Mike Halloran
Pembroke Pines, FL, USA

 

[foglght4]

Unfortunately, this is a system I can't change (oem production). I just got stuck with a calculation that hasn't been done in over a decade and going back through my old college books not finding much to help me.

The other awesome part is that the copper we are using has a varying yield strength between 65 and 330MPa and based on my simple bolt calculations (25nm bolt torque, 12mm fastener and the 0.2 condition factor) I get about 10.4 KN of force, which when I use the area of the copper gasket translates to about 125MPa of stress. So, it looks like in some cases it would easily crush properly, but if the yield was higher on some off batch, it wouldn't crush at all and potentially not seal.

I must be doing something wrong here converting the axial load into the flange clamp load.

 

[Tmoose]

Table UA-49.1 page 5.

http://www.seridium.com/downloads/Petrochemical_Gaskets_Handbook.pdf

soft copper seating stress 13,000 psi

 

[MikeHalloran]

It smells like you got this plum assignment because some of the washers are leaking. ... as your calculations suggest. Check again, and again, certainly.

Now it's time to retrieve some washers that leak, and some washers that don't leak, and try to estimate their respective yield points. You'll need to do a fair number to have statistically significant data, of course. Then you try to figure out how the purchasing spec got to be as wide as it is, and use your data to talk someone into narrowing the spec sufficiently to solve the problem within the narrow solution space to which you have been constrained.




Mike Halloran
Pembroke Pines, FL, USA

 

[EdStainless]

People make these to fit the bolt shank so that they center, but as you mentioned there is a radius there. They really should sit in a spot-faced recess and the washer should be narrower than the bolt head, that way you can generate lots of clamping pressure without risking breaking bolts.

Yep, I think that Mike is right, you are shoveling someone else BS.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[GregLocock]

I think an afternoon in the workshop with a torque wrench and a boxful of washers might be time well spent

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

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

 

[Tmoose]

I have an older no-name (but possibly US made) floor jack with a threaded plug with sealing copper sealing washer into the high pressure section.
The housing spot face was machined at an angle to the thread. Or maybe the thread was punched into the housing at an angle. I forget. Anyhow Some work with Dykem or probably a black sharpie revealed the severe angle mismatch, maybe 30% contact, and the intended mating surfaces were torn up to boot. At first I thought the head on the plug was crooked. By smoothing the surfaces and adding compliance (multiple softer washers) I was eventually able to get a fairly serviceable seal. If that had not worked I would have re-spot faces the housing, but first probably would have set the sealing washer in a bed of plastic steel type epoxy.

I'd be real interested in testing the contact patterns a few of the known leakers.

 

[foglght4]

The problem exists due to bolts being out of DIN spec and leakage occuring. Now we are being asked to revisit some calculations that haven't been done in a long time. While it is annoying, I gotta get it done. Just part of the job sometimes.

I am struggling with the yield strength. The copper is C11000. Yield strength seems to be all over the place from 69MPa to 375MPa. At the lower end I crush it, at the upper end I'm not even close to plastically deforming it. I have noted some washers come back with obvious plastic deformation, others do not, which would explain localized plastic deformation based on the plug having runout out of spec.

ugh

 

[Compositepro]

Anneal all your washers before using them. Copper work hardens, so the acts of rolling it into sheets and stamping out washers makes them hard. If the yield strength (hardness) is variable, then they have not been annealed. Annealing simply requires heating them above a certain temperature for at least a certain time. You can Google that.

 

[Tmoose]

"The problem exists due to bolts being out of DIN spec...."

Bolt head face runout to thread ( or vice versa ) or head flatness or finish or ???

This seems to suggest DIN copper sealing washers should be 45 HB maximum.

http://mdmetric.com/tech/DIN7603tdc.pdf

Just wondering.

regards,

Dan T

 

[sreid]

Is it possible that AN900 Crush Washers would solve your problem?

http://www.aircraftspruce.com/catalog/eppages/an900.php?clickkey=7239