I would like to get some idea for our trouble-shooting by discussing the stress analysis method of rotating equipment piping with members.

We have trouble in a turbine driven compressor piping. The coupling bolts between compressor and turbine shaft were broken as soon as start-up of turbine in few seconds. Of course, vendor and our mechanical team have checked all possible sources one by one. After long deliberation, vendor reported the most possible source of this failure is the excessive nozzle loads from external piping and they pointed out the friction loads. In my experience, I do not think the piping is a source of this trouble. Although piping loads are higher than allowable limit, these excessive loads will only cause higher vibration level than allowable limit, other than coupling failure.
We thoroughly checked the stress analysis again and the installation conditions such as piping supports, dimension, and weights at site. The deviations between original design basis and installation are minor in view of the stress analysis. The nozzle loads are well within NEMA allowable limit for normal operation condition by neglecting friction. We did not consider a frictional effect for normal operating condition and we used hot-setting -spring other than cold-setting for the spring supports adjacent nozzles. I would not like to discuss the setting method of spring since this may be controversial topic. Fortunately the differences of nozzle load between two setting methods are minor and both are within allowable limit.

The outstanding problem is friction loads. Turbine inlet and outlet piping(inlet 10”, 480 Celsius, outlet 14” 200 C) have spring supports(F-type) for first three points with Graphite sliding assembly on the top of spring load flange respectively and the rest are resting(shoe-type) support on beam. The result of Caesar II analysis by including friction factor(0.3) only for shoe(resting) support shows to exceed 200% allowable loads based on NEMA.
Our practices for the stress analysis of rotating equipment piping are as below;

1) Friction is considered as a short term transient load. Allowable load limits on equipment may be increased by a factor of 1.5 when considering normal loading plus friction loads.
In my opinion, generally the piping connected to turbine nozzle is fully hot with proper thermal expansion during compressor turbine warming-up. During this warming-up stage, the friction loads will be imposed to nozzle but not results in turbine failure, even though friction loads higher than 150%. Once thermal expansion occurs during warm-up, the friction effect is negligible during normal operation.

2) The rotating equipment allowable loads are the loads permissible for long term continuous operation. These loads are therefore compared with the piping loads calculated based on normal operating temperature. However unusual design or flexibility temperature conditions, such as steam-out, loss of fluid, seismic event, and so forth which occur mostly when the rotating equipment are not operating, estimated by simple temperature ratio, exceeds the allowable loads by more than 50%.

3) The nozzle loads may be calculated using the modulus of elasticity for the operating condition.

Based on above,
a) Are above our practices reasonable ?
b) Could the friction included loads really cause this failure ?
c) Is piping stress engineer responsible for ensuring the safety of turbine during warm-up condition ?

Our final intent is to show and ensure the external loads from piping did not result in this failure. Also we want to get out of trouble.

I hope that many members advise good experiences to us for this topic.


[ June 14, 2001: Message edited by: Soon Ryang, WEE ]

I usually do not deal with off shore requests but this was to hard to pass up.

1st did the F cans come with center guide posts? If they did not as long as the horizontal dispacements do not exceed the clearance between the column and the to plate there will be "no" load.

2cd Friction, I find your logic and opinion to be be ludicrous. What would you rather push a 400 Lb M on wheels or a 400 Lbm M sliding across concrete? Think about it sheesh. As for Duration being a factor, If I ask you to hold a 1000LB weight for only 1 second is that any better than 1 minute?

:rolleyes: John C. Luf

[ June 14, 2001: Message edited by: John C. Luf ]

I am not an expert in trouble shooting but a guess is that was there any steam bypass operation on this line during startup which may have to be checked.
Also you said that the failure occured within seconds of startup.
Again a guess is that can temp differentials cause such a failure in such a short duration of time?

Have you included the pressure effects in your analyses? Including the Bourdon pressure effect and less significantly the stress stiffening due to pressure (options in Special Execution Parameters) can significantly affect your nozzle loading, your pipe being 10" and 14" will produce significant pressure loads (ask a fireman!). These pressure effects can significantly increase your nozzle loads especially the moments even in the "cold" case (W+P1), and where the piping system is not suitably restrained by guides etc. as seems to be in your case. Because the failure is occurring "within seconds of start-up" before the piping system has "warmed-up", pressure effects are the more likely cause of the overload.
Regarding friction affects at supports, they MUST be included in any analyses especially where sensitive equipment is concerned, as mentioned by John Luf. I would carry out 2 sets of analyses, one including friction affects at supports and another set without friction. The real case will lie somewhere between these 2 extremes. If friction at supports causes nozzle overloads, place low friction pads at supports with the highest friction loads closer to the nozzle until the nozzle loads become acceptable.

I suggest that you become very familiar with API RP686 very quickly. This RP will assist your questioning as to the status of the equipment and piping prior to start up. Did you analyze what was built? Did they build what you analyzed? Was a check list used by operations prior to startup? Was a stress engineer present or was the system reviewed by a stress engineer?

There are many analysts who consider friction for the same reasons that you did (rationalized the effects of friction.) It certainly makes the stress analysis easier and at least the analysts relationship with the piping designers doesn't suffer. I consider friction in my analysis and it does result in "technical discussions" with others.

Friction has varying effects on piping and greater effects on equipment loading. That probably gives fuel to the analyst that minimizes friction on piping because his experience hasn't indicated major problems. How many large turbine and compressor combinations has the average analyst reviewed. In a life time, the number isn't that high. The more I read, the more I'm convinced that one should consider factors that effect the analysis.

I didn't see any information about the compressor, but the displacement of the compressor side of the coupling relative to the turbine side should have been considered. Again the API RP will add years of experience to your analysis. Learn the contents of the alignment section. Read the checklists, and ask if they were followed.

The problem resolution could come from something as simple as jack screws and equipment feet interference or a tack weld not removed from a support that was supposed to slide.

Somebody missed something (obviously) and the stress engineer is the primary target because we know that the calculation can't change as fast as a spring stop can be removed.

The percentage exceeding the allowable would most likely be accepted by the equipment manufacturer prior to contract.

I bet that next time you use friction and give the loads to the manufacturer even if you find the problem was due to travel stops not being removed.

I just had this confusion after reading Luf's remarks . I always felt that duration does matter.When a pipe is heated to higher temperature intially we have elastic strains coming into play and latter plastic strains .Ultimately bcos of temperature we have self limiting stresses induced.This condition is represented by expansion stress range term we come across in the analysis.For selflimiting stresses and consequently its associated terminal reactions in the final operating state it takes some time to attain .The final condition of the piping configuration is an equilibrium.Is time not an important factor to achive equilibrium.If yes then there is a transient state before the piping system achives equilibrium.Coming back to the turbine problem ...the heating during startup must be taking some time .This interveaning time must be enough for the piping to achieve the equilibrium i,e the final state.Now coming to friction ..it comes to play during the time the system is tending towards equilibrium and its role ends after the system reaches equilibrium.So in the final equilibrium state the forces at the terminal points are due to the stiffness of the whole piping system.Does that mean that during the transient state the terminal loads are a combination of that due to stiffness and friction , both which are varying in magnitude as the piping system tends towards equilibrium.The friction load is decreasing and the stiffness loads increasing plus at the same time selflimiting itself.Going by these thoughts of mine the loads experienced by turbine must be less during startup than during final operating state.

I will post what I hope is a final set of responses to what I feel are some basic concepts in piping design and analysis around sensitive ROTATING equipment.

1) Friction should be considered carefully. The designer must use friction ONLY as a conservative factor to the design and never as an aid. Friction is inconsistent due to a wide variety of factors even when TFE or other slide bearings are used.

2)Rotating equipment loads are never separated into load cases like code stresses are. There is no such thing as far as a self limiting load to a piece of rotating equipment! It is not a simple lump of static wrought metal like an elbow (you gotta be kidding me either the turbines in Singapore are real robust or God is taking care of you!) The amount of time a piece of rotating equipment is overloaded is not a germane point of discussion. In a few seconds alignment can knocked askew etc.

3)Is friction the cause of this failure.... is it only overloaded 200%? Nobody in this forum is qualified to say (including myself).
a) Is the CAESAR II model correct in all aspects?
b) Were supports installed correctly?
c) Was the piping installed correctly and brought to the flange of the turbine?
d) Was the turbine properly installed, aligned, grouted etc.?
e) Was the turbine started properly?

4)Due to the rather large magnitude failure I strongly suspect that the overload is more than 200%, but in a court of law an expert witness can be called to testify that friction was not considered properly in the model.

End of story Check and Mate!
:rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes:

[ July 06, 2001: Message edited by: John C. Luf ]

While I certainly buy the argument that a nozzle that is excessively overloaded may cause the failure, I can't say I'm convinced in this case.

I've reviewed a number of existing systems for clients who, after putting up with years of high maintenance issues, realized that the piping configuration could be the issue. It is a a surprisingly major leap for most people not involved in stress analysis to think that a piping configuration can have anything to do with the reliability of rotating equipment. But's that another topic altogether

Now, there are three reasons I doubt that the piping system is the cause here:

1. Some of the systems I've reviewed have turned out to be 500% and more over published allowables during operation. Even under these loads, the machines don't fail instantaneously, but they do have to come down every 3-6 months for major work.

2. If the first support was so heavily loaded as to create a friction anchor in front of the pump, I can't see how they ever would have gotten it bolted up in the first place. The reason is that, unless the support were to be able to act directly under the cg of the pipe span, that much vertical load would impose a substantial moment on the connection at bolt up and I can't see the shaft passing the alignment check with the kind of load we would have to be talking about.

3. In order to have friction, something has to try to move. In the case of the systems we look at, that movement comes from thermal expansion, which is not an instantaneous process. As the lines comes up in temperature, it is slowly going to try to expand. The static friction load is going to build from zero, over time, up to the maximum that the normal force can sustain. I find it hard to believe that you would have a combination of such high normal force and quick expansion to produce a reaction capable of causeing such a catastrophic failure in a matter of seconds.

Of course, these are just the ramblings of one analyst. I know the question was specifically about friction and that, supposedly, the vendor had eliminated other possible failure sources. However, what is stated and what is true aren't always the same thing.

Now, my observations are based on my own experience, which is by no means exhaustive. I would certainly appreciate responses from the group of experiences where a piping system was so poorly layed out as to cause a instantaneous catastrophic failure of a rotating piece of equipement. I'd also be curious to know how it passed the shaft alignment check.

I saw that including the effects of pressure (Burdon pressure effect) increase the loads on the nozzles. If you didn't consider this effect in your analyze I think the loads will exceed the allowable with more then 200%.
I also think that including the friction in analyze is very important, but also the results must be interpreted to see if they could be real. In such cases (sensible equipment) I am using expansion joints and/or very flexible piping traces. I usually avoid the friction supports in that case because it is difficult to appreciate if the friction value is real, or it is to high. In any case, taking care about friction is conservative.

But, in my opinion, only these loads couldn't cause the failure "in a few seconds from start-up" (purely if these loads are not extraordinary high) . You have to check if the installation of the machine was correct (alignment, clamping by screwing up bolts and nuts, etc), if the operations indicated by the vendor was respected, etc.

In any case, it is hard to convince the vendor or anyone that the cause of failure is not the piping system, because of these omissions from the stress analyze.

I should like to know if you would discover the real causes of the failure.

Best regards,
Dragos

Are there any expansion joints in the system ?

A poorly restrained expansion joint could very quickly subject the equipment to excessive forces when the system is started up.