friction factor used for shoe slides 4

[pipesnpumps]

What does everyone use as the friction factor for steel pipe shoe on steel angle/beam?

I've always used 0.4, but a structural sub is designing the pipe support steel for a friend, and wants him to use 0.75.

Obviously, the 0.75 assumption is more conservative. Has anyone been bitten by going with 0.4?

The lateral loading on the supporting steel due to friction is way high at 0.75

 

[JohnBreen]

Hello,

In one form or fashion this is a frequently asked question on this and other discussion boards. The real answer IS NOT a numerical one. The reason I say this is because it depends upon why you are doing the analyses. If you are doing the analyses to get nice numbers for the owner to show the representative of the pump company then by all means use the 0.4 number (I mean, after all, isn't it in Mark's Handbook?). That number will also result in the pipe stress analyst being comfortable.

In fact, the first issue to consider is what point in the service life of the system you want your analysis to approximate. If the system (including the nice new, smooth steel plates (slide plate and base plates) is installed with excellent workmanship and an appropriate amount of the lubricant best suited to the application, the 0.4 number might be a good approximation (for some initial finite period of time). From day one of service, the coefficient of friction will degrade (get higher). Contaminants will get into the slide/base plate interface. The lubricant will deteriorate due to oxygen and UV (light) the steel will oxidize and some other dynamic effects might change the plate surface smoothness. So, then you can expect your system loads to increase as the "actual" friction coefficient goes to .6 and on to 1.3 and....

In a computer model, the initial STATIC friction is typically modeled as a spring that will be compressed when the slightest axial force is calculated - and the pipe will smoothly move with the first inch/ounce that is calculated. The thing about using static coefficients of friction is that they can be high enough, at start-up, to initially preclude pipe movement (do you pre-load the spring?). The compressive elasticity of the pipe will allow some compressive strain to "accumulate" (think of it as an axial "spring&quot The axial force along the pipe will slowly increase until the force can overcome the initial moment of (friction enhanced) inertia. Then there will be a movement. But as soon as the axial force is "released" by the accommodating movement the movement will cease. The movement will, in fact be discontinuous. As the slide/base plates corrode and as the lubricant degrades, the movement will become more erratic and non-linear. The point might come when the pipe moves in a series of "hops", and when (if) this happens, the smoothness of the plates will be affected and consequently the friction will increase. The point to understand is that with a series of these sliding supports in place, the response to axial force is non-linear and inelastic - even assuming that "perfect" and consistent workmanship leave you with the same condition at each support.

There was some testing done that was sponsored by an (US) National Standards Committee. The conclusion was that the initial (at time of installation) equivalent static friction coefficients vary according to design and workmanship. They never saw an equivalent static friction coefficient as low as 0.4 and they did see equivalent static friction coefficients in the range of 0.6 to 1.3 before the slide plate started "hopping". They concluded that there were so many variables possible with what was/is accepted as "standard industry practice" that any number representing the equivalent static friction coefficient would have to be developed by testing at the site of installation using the actual craftsmen before the using the number in analysis could be justified. In the early days of constructing nuclear power plants, that was exactly what was required.

From a practical point of view you have to come up with "a number" (or more likely, "numbers&quot. The best that you can realistically report from your analyses is that the loads on the strain sensitive terminal equipment will be within a calculated "range" (envelope). To report results more exact than this would be disingenuous. Do I think that it would ever be appropriate to use an equivalent static friction coefficient of 0.75 at the sliding support? Yes, if you are lucky. But as soon as your rotating equipment monitoring program begins to indicate increasing machinery vibrations, go get the grease gun.

Regards, John.

 

[JohnBreen]

Hello,

I suppose that I should mention that sometimes sliding supports are used with expansion joint (either bellows or the telescoping types).

There is an interesting analogy here. The bellows allows a continuous axial movement (initial preload and spring coefficients are appropriate in the computer model). Sometimes, in the telescoping (or sliding) types of EJ's the gland packing (seals) at the interface of the sliding cylinders is kept very tight to preclude leaking. Often these types of EJ's will result in the pipe moving axially in a discontinuous manner as the friction in the packing must be overcome by the axial force before the pipe moves (similar to the movement caused by friction at the sliding supports). It is not uncommon to hear the "popping" of the discontinuous movements in the EJ's (especially in a stem tunnel where there is no room for anything but these EJ')

The halting movements typical of these systems (and maybe the sliding guides also contribute) results in impact loading on the system "anchors". It is not unusual to find damaged anchors and broken anchor bolts in these applications due to the series of "impact" loads that are applied (in opposite directions) on "heat-up" and "cool-down". Over a period of years the maintenance people keep tightening the glands (seals) as they begin to leak (due to normal wear) and so the high friction is perpetuated during the systems operating life.

Something else to think about when strain sensitive terminal (maybe rotating) equipment is involved.

Regards, John.

 

[stressone]

Sir,

in our engineering company (Europe based)following friction coefficients are used :

- steel to steel : 0.3
- PTFE to steel : 0.1
- lubricated PTFE to steel : 0.05

These values have been aproved by several government agencies that are in charge of approving design and calculations. Sometimes we even use a friction coeff. of zero at the first support after rotating equipment, in order to simulate vibration.

 

[cceballos]

Hello, I agree with the values given by "Mr. Stressone", but when you say that those values have been aproved by several goverment agencies, could you inform me about any document, instruction or standard approving them?

Do you know any literature about this issue?

 

[stressone]

There is no specific code or standard that specifies these values.

The determination of the friction coeff. is based on tests.
Normally we use this value (0,3), present it to the client and ask for his approval.

As for literature : a (Dutch) book "Staalconstructies" (Steel constructions) mentions this topic.

Also Dutch Code NEN 6772 - Steel Structures gives some additional info., but no specific values.