hi!

I have questions regarding the use of NEMA SM-23 and API 617 of CAESAR. Sorry if this seems to be trivial.

1. Resolution point is required in this sheet. Is this the same resolution point as defined in API 610 (that is intersection point between shaft and nozzle centerlines) Also, in sample problem in NEMA SM-23, there is no resolution point mentioned to complete nozzle evaluation.

2. Sometimes, we can use 1.85 times allowables of NEMA for centrifugal compressors. Does this mean I can use the sheet of NEMA SM 23 to evaluate nozzle loads of centrifugal compressors (just input 1.85 as factor of allowables)?



Needing help.

Thank you in advance

The resolution point for a two nozzle machine is usually the centrline and face intercept of the largest nozzle, although it is often not clearly defined. For a four nozzle (compressor) it is usually the shaft centreline, but that's not written in any Code.

A so-called NEMA factor of anything up to 10 (largest I've seen) may be applied to SM-23 allowables when dealing with compressors, but 1.85 or 3 times are the most common. Normally it's 1.0 for steam turbines.

I strongly recommend contacting the machine vendor early in the project to agree both the 'factor' and the resolution point. In addition you should agree the 'Dc' value .... usually it's based on nominal diameter, but could be based on ID's.

hi all, i'm a new comer here.

in addition, is correct that the resolution point is also affected by global axes direction?

for example, +Y direction is from Suction Nozzle to Discharge Nozzle. Suction Nozzle is located 500mm (behind) Discharge nozzle, and we choose Suction nozzle as the resolution point.

thus, it means that we need to put (Y) = -500 for resolution point of Discharge nozzle. is it correct?


best regards,
chigurh

@mister moverz

what i know, Dc is combined diameter of suction & discharge nozzle. we use this if want to analyze suction&discharge nozzle together with respect to C.G. of the compressor.

1st, calculate De (equivalent diamter of both nozzles);

De = (Ds^2+Dd^2)^0.5
Ds = suct. nozzle diamter
Dd = disch. nozzle diameter

if De > 9in, than we need use correction factor:
Dc = (18+De)/3 (in)


similarly with individual nozzle analyses, we use Dn (nominal diameter). if its size is above 8in, than we use correction factor:

De = (16+Dn)/3 (in)


CMIIW master, please

Anton,

Whatever direction your compressor is located, you need to use the NEMA axis system, as shown in the SM-23 document.

Why are you referring to the CG of the compressor ? That has no relevance to a nozzle load evaluation.

Regarding nozzle diameter, it is not clearly defined in the Code and nominal is normally used. However I have seen the inside diameter referenced. Like I've said, it should be clarified with the vendor.

It seems that I've resurrected an old thread. I am having a great deal of difficulty getting some preliminary modeling to pass the NEMA allowables. The main reason is that NEMA makes no provision for class or schedule. Many times the main steam line will be CL2500 with sch 160 piping. This system is incredibly stiff in addition to usually having a forced displacement at the nozzle connection. Getting my sustained case and thermal case to pass is a pain (dare I say, impossible) let alone the occasional cases. This is with an SIF of 1. NEMA allowables seem ridiculously conservative even compared to simple API pump allowables. Can you elaborate on this NEMA factor? I must of missed it...

The casing of a steam turbine, or a power recovery turbine in an FCC plant for instance, tends to be quite thin, and very hot in the FCC case, consequently local casing deformation and blade contact are a real risk with excessive nozzle loads.

Regarding a NEMA limited compressor, the casing is usually very robust and the weak point is commonly the frame or plinth. I have seen 10x NEMA on a compressor that was 420 bar outlet pressure and had a casing around 6 inches thick. The nozzles were machined directly into the barrel, so it didn't actually have mating flanges as such. Clearly very different to a steam turbine.

To get your S160 pipe nzzle loads down, have you considered routing the pipe around the turbine to cross both machine axes, then applying line stops on the axes ? This should cancel out local expansion, reduce direct forces and consequent moments to secondary effects and might get within allowable limits.

Thanks for the response. The task that I've been assigned is to layout several turbine piping configuration that can be tied back to a 6 way restraint (anchor) so we have a modular system that we know will work for future projects.
I've anchored my system at the point of zero growth so for sustained and thermal cases, this works. BTW, my NEMA allowable is 1x allowable (its a steam turbine). This work in theory but the system is not very robust and even small dimensional changes impact the loading by a large amount. Although, I may have to run a flexibility analysis on the nozzle to get some sort of stiffness which would make the nozzle loadings much less sensitive. I would hope the actual vendor would have the option to build the nozzle / casing to satisfy x2 NEMA allowables. That may be wishful thinking but it would sure be nice.

I have seen a 20 x NEMA compressor. Not a fisherman story .

A Nuovo Pignone compressor. I think they are famous for this aspect.

Regards,

What about the following two solutions...?

I am not stating that it will be technically correct but just wanted to pass a point here and further improve my knowledge at the same time.

1. If it is a high pressured system, and there are break out flanges just above the nozzle, why not select one of those grayloc/techloc hub clamp flanges. It will save space and reduce your flange weight definitely. Sustain loadings will be reduced and eventually assist with the thermal loadings.

2. If there is a reducer involved, try to be place it a slight distance away from your nozzle and not at the nozzle. Most of the time there is a reduction in size of the piping. Similiar to a pump system. This may help with the reduction in sustained loadings.

Once again, this is just my opinion and I am open to any criticism in order to learn.

Cheers.

Sustained loads are not the problem so far as compressor nozzles and similar are concerned. Normal practice is to ensure these loads are zero by careful alignment and spring support balancing. The mating flange will usually be required to be practically perfectly aligned when the bolts are removed.

Consequently, loads to be compared with NEMA / API617 tec. should be thermal and displacement in nature.

Thus, I would say that use of Graylocs will not help nozzle loads. However, a length of smaller size pipe local to the nozzle, rather than a reducer actually on the nozzle should have a significantly beneficial effect. If you can get it past the process engineer.

@MoverZ: Practically speaking,zero loads at mating flanges are not possible for sustained condition because changing spring parameters to bring loads close to zero plays bad with Flange Alignment or with thermal loads.

I've seen compressor vendors who were adamant on zero loads or zero displacement at mating flanges. It was really tough for us to make them understand that this was not possible.

As far as alignment is considered, fit-up spools are the best option to reach a perfect alignment.

In my experience, careful use of spring setting and control of pipe position through use of adjustable supports will bring the nozzle alignment to within the vendor's allowed limit. That would be considered near enough to zero loads in the real world.

I would not recommend fit up spools since compressor piping is often heavy wall. Welding then heat treatment may cause significant residual bending, thus defeating the object.

My use of the term ... zero loads ... applies to the cold, installed case. The inference being that specifically a NEMA balance stress calc does not need to include weight effects. That calculation may in any case be unique since it could well be based on operating rather than design thermal conditions and hot Young's modulus.

Hence, would you fellow engineers out there agree with the following approach to determine minimum displacements of the mating flange at the nozzle during the installation with the springs 'unlocked' to ensure allignment is met.

1. Once you have painstakingly brought the compressor nozzle loads down to within the allowable limits, save the CII file under say "flange disconnected"
2. Remove thermal displacements at the compressor nozzle or modelled anchor if compressor vendor has not provided any. Flanges would be considered disconnected in this scenario.
3. Simulate your springs in CII to be 'un-locked'.
4. Run the analysis and check only the sustain case displacements.
5. Ensure displacements at the flange when disconnected to be within +/- 3mm in lateral and vertical directions. 3mm is a thumb suck figure...

Once again, this is just my opinion and I am open to any criticism in order to learn.

Cheers

@MoverZ:I beg to differ but using adjustable support (I hope you mean a support arrangement with jack screw to lower or raise the connected pipe) for alignment for bigger line sizes (say for a 42" compressor line) at site is practically very difficult. I was involved in a detailed fit-up for a 42" compressor line & synchorised efforts of Stress & Construction yielded in a perfect alignment with literally no troubles.

May be my case was one of the lucky ones ...


I fully agree with you that Compressors & Turbines shouldn't be checked for sustained case loadings as per Nema.

@PiperTan: One correction for point# 4&5 -

"Run the analysis and check only the WNC case displacements"

Alignment limits vary from vendor to vendor.I know a few vendors who don't accept 3 mm misalignment. So this figure can't be generalised.

Sorry mate, I forgot to mention to confirm from Construction whether line is to be insulated after alignment is made or is it partially insulated (leaving only one or two spools uninsulated in vicinity of nozzle) or alignment is done with line being insulated fully.Then this needs to be incorporated in alignment check file.
Status of Insulation plays very vital role in alignment of nozzles with higher sizes!!

Remember that the weight displacements and loads you get from Caesar are usually based on nominal pipe wall. This can vary by +/-12.5% so the data you get may vary by 25%, and stiffness will similarly be affected.

Keep the analysis realistic and don't expect too much of calculated results based on inaccurate input data.

I agree with MoverZ on the allignment point..

basically, compressor deals with gas so fluid density is not high.. So first, disconnect the flange from compressor (i.e remove displacements). and run the system.. so all dead weight will come on the first spring.

From Output, consider WNC loads to design the spring and adjust the load if required..

It will help in allignment issue also..

Correct me if i am wrong...