I am doing a live steam line. Based on the same arrangement and restraints, without the friction (keep the Mu blank)the moments and forces to the nozzle of turbine are within the allowable values required by the Turbine manufacture; if put 0.1 (Teflon on Teflon or rigid hanger), then the moments and forces to the nozzle are totally changed.

Any suggestions?

Thanks in advance.

Best Regards,

Shuang

Dear Shuang,

in my opinion both cases (without friction and friction included) should be analysed, and the worst results should be used to compare them with the allowable nozzle loads. Friction always will be present in the piping system, the exact value (0.3 or less(or even more for older pipe supports?!] is often subject to discussion.
But especially in the vicinity of rotating equipment or where piping systems with low eigenfrequencies are to be investigated the sometimes helpful effects of friction cannot be regarded as always available. (A support that lifts up for a short time due to vibration of the pipes won't transfer any friction). So my intention is always to look for both results - the true nozzle loads might be somewhere between these values.

Hope this helps a bit and all the best,

Veit Bockemuehl
engineering services network gmbh
Hamburg/Germany
vb@esn-gmbh.com

Additionally, please read the section in the Technical Reference Guide on friction. The discussion begins on P6-16. The conclusion to this discussion is that there are <em>multiple equilibrium solutions</em> once you include friction in a system. So evaluate your results carefully.

As was mentioned above, the value for the friction coefficient is subject to discussion (or disagreement). You should always look at several cases; no friction, a low value for "mu", and a high value for "mu". You can easily do this now with Version 4.40, since you can apply a global friction multiplier on a <em>load case by load cases</em> basis. (If you haven't found this yet, check out the Load Case Options tab of the Static Load Case Editor.)

Please ensure that GAPS are included in the restraints along with friction (as suggested by Richard Ay).
It is also important to distinguish whether the gaps are standard or special; e.g. zero gap restraints need a special detail from design and installation point of view and turbine nozzles are generally sensitive to gap distances because the allowables are quite low. In other words, in compressor analyses, a series of sensitivity checks are required.

Nimal Jayaratne
Lead Pipe Stress Engineer
TIGA JV-Bayu Undan Gas Recylce Project
Perth WA6153
Australia
e-mail: njayaratne@tigajv.com.au

Hello,

One should always run a system(especially a critical system connected with a turbine, pumps, compressors,Air fin cooler) with friction & without friction.As we know, friction effect is transient.Sometimes friction helps in reducing the Nozzle loads.So one should not always understand that with friction, results are conservative.

Regards,

Saneh Gupta
Stress Analyst,
Fluor India,
N.Delhi

Friction effect modelling is still a matter of discussion in Pipe stress programmes. Because of the non determinictic nature of the problem, the method suggested by Sobieczanski is more or less followed in Pipe stress programmes which is a numerically iterative treatment of this problem. It is a known fact that during the "movement phase" when coefficient of kinetic friction comes into play, the equations of static equilibrium are no longer valid.Hence the effect of Friction modelling is just to give the user a feel of the effect of friction .Suppose your load is 100 Lbs without friction and it is 120 Lbs with friction.This does not mean that there is truly an increase of 20 Lbs due to friction effect and will "disqualify" the system..It is just a feel that the load might increase due to friction.Even in Caesar if you change the friction stiffness( even within limit as suggested in Caesar's technical ref. manual) your system can change its status from a " qualified " one to a "disqualified" one.

So till date in absence of a more sophisticated deterministic approach to solve this friction problem the effect of Friction need not be necessarily given a very serious consideration ( unless for very high temp.or very big size line) and the results are best treated by non friction analysis.Many Caesar II and other programme users may strongly disagree with me.However for very high temp. and critical services it is better to resort to " non friction supports like STRUTS , RIGID HANGETRS etc. to avoid the effect of friction ( at least close to the equipment) .

A. Bhattachrya

Stress analyst

Bechtel Corporation

Friction loads can be highly variable for several reasons. The allowable loads for a turbine might be extremely low for smaller equipment - a SIEMENS 3500 kW turbine had internal clearances of 0.5 mm and there was much concern for piping loads deforming the casing. So what can be done ? The effects of friction should be minimized by the selection of types of supports with low or no friction, i.e. rod hangers, spring hangers, slide plates with low friction material, and expansion joints (tied universal joints are my preference).
Piping does not behave in a linear manner with friction, liftoff, and gaps at guides / stops. The reference cited by Saneh Gupta has data showing the friction variation with velocity. This is important because usually the heating occurs fairly rapidly - piping support has more velocity - while cool down occurs slowly due to insulation. The last interval of cool down has a smaller temperature delta and a pipe support has a very small velocity - almost static coefficent of friction occurs. The piping supports on bare steel will have "stick-slip" friction behavior at the low velocity, so the actual coeffient of friction is difficult to define.
Turbine startup / shutdown sequences can also result in rapid temperature changes with extreme piping movements. SIEMENS vendor data for nozzle displacements gave normal and "no-load" thermal displacements, that showed there was greater displacement (temperature) during the no-load time of startup / shutdown. During the no-load condition there is little resistance to steam flow and very little energy removed from the inlet steam, so exhaust temperatures are hotter - almost at inlet steam temperature. The turbine exhaust piping could see very rapid temperature increase, and slower return to normal condition. An expansion joint could be included in exhaust piping to allow these variations and movements.
Guides and stops can be a major souce of variation in friction loads. If a combined guide and stop support has say 0.5" gap east, and 2" gap north, then it is probable that the 0.5" gap will close first, and friction load prevent the 2" north movement going to full 2" travel for loads to balance out. After expending much effort to remove friction from the piping system it is necessary to add guides and stops into the design to limit the piping movements during a "steamhammer" event due to turbine trip valve sudden closure. When trying to control the loads on a turbine, I find it easier to think of controlling the movements at supports, so that the piping forces are in directions to balance resultant loads on the turbine. Due to friction, ordinary piping might not return to its original position. It could be likened to an instrument control loop with linearity, hysteresis, and deadband problems. Repeatibility and accuracy will not occur without careful design of a system. One old engineer even used the Code conservative material allowable stress to avoid plastic yielding of pipe and its self springing that would cause unpredictable loads on turbine.