SETTLE A FRICTION ARGUMENT? 2

[FDS2008]

I have a physics question that I can't seem to answer using textbooks.

I have a vertical pipe that will be clamped to some structural steel. We are trying to calculate the forces required to restraint the pipe based on friction. The clamp will pull the pipe towards the structural base and there will be two different contact surfaces.
1 - The clamp
2 - THe structural steel frame

The two possibilities that we are trying to settle in the design camp are as follows. The first is that the friction force will be doubled (i.e., friction on item 1 + reaction on item 2 = 2 x friction force) since there are two friction surfaces. The second one considers the friction force on the structural steel frame as a reaction of the clamping force and the pipe would opnly need to overcome the friction force of either surface for it to slip axially. This would mean that the resistance is equivalent to the clamping force only.

What do you all think?

 

[rb1957]

i think it's either/or not both ... one interface or the other will slip independent of the other, no?

 

[PeterStock]

I agree, once one of the interfaces starts to slip, the load on the other goes to zero.

Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

http://www.infotechpr.net

http://www.linkedin.com/profile/view?id=30064526&trk=tab_pro

 

[zekeman]

I see 3 surfaces, not 2

clamp to pipe
pipe to frame
clamp to frame

The weakest friction interface will let go first.

 

[zekeman]

On second thought, the clamp to frame friction would not be involved in this debate.

 

[zekeman]

On 3rd thought, I think that so long as the friction (frame to clamp or pipe to clamp) can support the clamp weight, the friction between the pipe and frame is the only criterion for holding the pipe.
I hope this is my last comment on this simple but elusive question

 

[ivymike]

no diagram provided, but assuming that we're talking about a round pipe sitting in a saddle with a u-bolt clamping it to the saddle, and that the saddle is rigidly fixed to the wall, then...

two interfaces of concern:
1 - The clamp - to - pipe interface
2 - THe structural steel frame (saddle) - to - pipe interface

friction at each interface must be overcome to have slippage at the interface in question. Slippage of the pipe w/r/t the steel frame can happen without slippage of the pipe w/r/t the clamp, assuming that the clamp can deflect. I would anticipate this behavior, in fact, since the clamp definitely can deflect, and would in many instances increase its clamp force as it deflected, thus attaching itself more firmly to the surface of the pipe (still a failure to restrain).

Unless the clamp is intended to support a large fraction of the supported load via shear and bending of the clamp itself, I would recommend designing the part so that the friction at the supporting structure is more than sufficient to support the full load. This will avoid axial loading of the clamp (which I assume you want to avoid).

 

[Tmoose]

I think each surface provides its own frictional resistance. How does one surface "know" that the other exists? Multi plate clutches handle more torque at the same spring pressure.
Despite the basic physics premise that friction is independent of area, i'm guessing the clamp wrapping part way round the pipe will make the greatest contribution, unless there is a saddle at the structural steel as described by ivymike.

My attempt at a first cut would be to consider it half a "riser clamp"

http://www.pipingtech.com/products/ptpcat/support/clamp/fig90.htm

Note for full clamping force the clamp can not "bottom" on the structural steel after tightening.

http://www.webbikeworld.com/Motorcycle-tires/avon-storm/fjr-1300-front-pinch-bolts.jpg

 

[rb1957]

but they wouldn't act in parallel as the op suggests "the friction force will be doubled".

they are separate and one or the other will slip

 

[rmw]

Is this a schoolwork problem?

rmw

 

[chicopee]

Draw us a sketch because I am a little confused with your description.

 

[IFRs]

The pipe does not know that one side is a clamp and one side is a beam, and it does not care. In a simplified, ideal world, wherever it feels contact pressure there will be friction. It's up to you to decide how ideal your situation is. For instance, if the surfaces do not stay parallel then secondary issues arise.

 

[WillCole]

I am not sure I understand the description 100% but..

There is no 'doubling up' of clamping force (if that is what you mean) because there are two surfaces. You have increased contact area and so increased resistance to movement, but it is a function of area.

 

[Tmoose]

Hi rb1957.

This is how I pictured his system, even after re-reading it a few times.

http://www.ted-kyte.com/3D/Pictures/U-Bolt-MufflerClamp.jpg

the flat bar represent the surface of the structural steel. The U-bolt is the clamp.

There will be some friction available at the flat bar/pipe interface as a result of the force applied by the clamp.
There will be some friction available at the pipe/u-bolt interface due to the same force being applied by the clamp.
To slide the pipe axially one force greater than the sum would have to be applied.

If, by chance the clamp friction and the steel frame friction are the same the force required to slide the pipe will be 2X the force required if the clamp were superTeflon (coefficient of friction = zero).

"Adding plates to a clutch unit to form a multi-plate clutch will increase its torque capacity, without increasing spring strength or clutch diameter."

http://www.blu808.com/assets/images/50057-350.jpg

If the question was whether the steel surface friction and the clamp surface friction should each be doubled, then I was mistaken.

Dan T

 

[gruntguru]

Draw a FBD and set the pipe in motion. The two contact surfaces will both contribute friction and the contributions must be added. For a rigid clamp the static friction contributions must likewise be added. A flexible clamp adds complication as mentioned previously.

It is not a function of area either - just normal force and friction coefficient.

Engineering is the art of creating things you need, from things you can get.

 

[zekeman]

I took another look at this and come to the following conclusions

1 So long as the pipe does not slip, it is impossible to allocate the distribution of friction but the sum of the 2 is surely the longitudinal force on the pipe.

2 As the longitudinal force increases, slippage will occur at both interfaces simultaneously at a force equal to the sum of the static friction forces at the 2 interfaces. My reasoning is that you can't have one interface slipping and the other static, since the very act of slippage must be inclusive.

3 Looking at 2, it may be possible that the distribution of friction at forces less than the limit may be proportional to the static coefficients, but that is a gut feel, since it would neatly fit the simultaneous slippage, I postulated.

 

[ione]

You have two contact areas. Let’s assume each one is characterized by its own fiction factor (µi). Each contact area represents a boundary. Compute the normal forces acting on each surface (Ni). Friction is independent of area. Compute the product of each friction factor by the relevant normal force. The higher µi*Ni will rule the motion.

 

[rb1957]

some of the posts have made me think (!?) about static friction vs dynamic friction ... for static friction i think the lower frcition interface will rule, but maybe under dynamic friction, ie if both interfaces are sliding, the total friction is the sum of the two interfaces ...

 

[corus]

Friction has nothing to do with area, but with the reaction force being applied. Calculate that force and you have the friction force resisting any applied loads.

Tara

 

[ione]

rb1957,
The only difference between static and dynamic (kinetic) fiction is that the dynamic friction factor is lower than the static one. When you equal or overcome the higher µi*Ni (being µi the static friction) motion will happen.