Thermal Expansion Of A Large AC Motor 3

[desertfox]

Hi

Recently we received details of a large AC motor which will be mounted on a fabricated skid with 4 bolts M56, however about a week later we were informed that the bolts need to be increased in size to M64 because of thermal expansion of the motor casing.
Now this is where I don't understand the logic, I understand if a motor driving a pump gets hot then the motor shaft can expand vertically (mounting foot to shaft centreline) and cause misalignment but axial displacement is normally allowed for in the couplings and motor bearing clearances.
I can't see how increasing the foundation bolt preload would help with the vertical deflection of the motor shaft and in addition they want us to fit four taper pins through the motor feet after alignment, so even if we fit larger bolts any axial force on the base will be taken by the pins, so all the increased bolt preload just serves to increase friction between mount pad and skid.
I have done a Google search but cannot find anything about increasing foundation bolt size and how it might help with restricting thermal movement, anyone got any similar experiences to share?

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[7anoter4]

I am not sure, but I think it is connected to the resonance phenomenon. If the axial distance between
supports increases the resonance rpm decreases and may be close to the rated .In this case the vibration forces acting on the bolts are bigger.

 

[desertfox]

Hi 7anoter4
Thanks for the response!

Nobody as yet as mentioned resonance rpm in connection with the holding down bolts and to my knowledge the axial distance between supports hasn't changed.

Originally the bolts were M56 specified by the motor manufacturer and those bolts were based on the motor having an internal short, which would produce about 500kN in each bolt, I checked the figures myself and the M56 were adequate for the 500kN and the forces produced by motor torque are very small when compared with the 500kN.

I have asked the manufacturer for an explanation because I don't think this bolt size change is anything to do with thermal expansion or even fatigue for that matter but I would like to understand more about the bolt sizes relating to motor holding down bolts before I get a response back from the motor people.

anyway thanks again

desertfox

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[electricpete]

It is generally not desirable for the machinery to move away from the aligned position.
Misalignment (thermal or otherwise) is accomodated by flexible coupling as you say but will still create reaction forces that may tend to move the machinery.
Higher hold-down bolt clamping force will help prevent undesired movement which disturbs the intended alignment.

Dowels tend to be viewed more as assembly aides (to get back to the same target) than as means to prevent movement as argued by Dan Timberlake in this thread:

http://www.maintenance.org/topic/dowelling-is-necessary

(I argued the opposite side for awhile, but I defer to his expertise)


=====================================
(2B)+(2B)' ?

 

[desertfox]

Hi electricpete

Thanks for the contribution and the link you left.

I agree with your statement about higher bolt preloads reducing the chances of any movement due to increased friction, however imagine even with the higher bolt load the bolt shanks might not be engaged with the edge of any of the holes, under those circumstances it's the taper pins that will prevent any further movement.

Don't get me wrong I believe the taper pins are there to maintain the original alignment after the motor as been installed and positioned.
The original bolt size was M56 and that came from the supplier,I checked the bolt size myself and agreed based on the forces generated from the motor under fault conditions.

To date there as been no calculation given from the supplier or justification for the increase in bolt size, now this makes me suspicious because I was told the bolt increase was due to thermal expansion, so if they know they had to use an M64 instead of an M56, then someone must have done a calculation and must also know what the bolt preload required is, under those circumstances the bolt preload should not take over a working week to be passed on.

All that said Pete and from what I have read the biggest problem is thermal growth from the motor shaft in the vertical direction and I cannot see how increasing the motor foot bolts will significantly prevent this.

Regards

Desertfox

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[electricpete]

I agree with your statement about higher bolt preloads reducing the chances of any movement due to increased friction, however imagine even with the higher bolt load the bolt shanks might not be engaged with the edge of any of the holes, under those circumstances it's the taper pins that will prevent any further movement.

I'd suggest that the bolts don't prevent movement by contact with the sides of the holes, rather they prevent movement by (vertical) clamping force which increases the friction you mentioned which (in turn) resists horizontal movement. The larger bolts will usually permit larger clamping forces to be applied.

You'd think the taper pins (dowels) alone would be enough to prevent movement, but again I'd refer you to Dan Timberlake's comments to the contrary in the linked thread (Dan goes by the handle Tmoose on eng-tips).

All that said Pete and from what I have read the biggest problem is thermal growth from the motor shaft in the vertical direction and I cannot see how increasing the motor foot bolts will significantly prevent this.

I can come up with a scenario that sort of fits the need for larger bolts, but no guaranteeing that it's right.

Higher bolt load will not prevent vertical movement, but should prevent the machine from moving in response to the static and dynamic reaction forces from the expected misalignment which results from vertical movement. Note that even if a thermal correction is applied during alignment (misalign the machine at cold conditions in order to make it grow into misalignment at hot operating conditions), it will still be operating misaligned until it warms up.

Now another question might be what does vertical growth have to do with horizontal movement. I wouldn't intuitively think they are related, but the reaction forces are complicated and in fact I believe vertical misalignment can result in horizontal reaction forces. If we use the model on page 2 here:

http://turbolab.tamu.edu/proc/turboproc/T5/T5pg111-116.pdf

Assume Pure vertical misalignment => delta-Y and Phi are non-zero (delta-X and theta are zero).

By inspection of the equations we can also see:
MY1 is non-zero.
MY2 is non-zero
FX1 and FX2 are non-zero. These are the horizontal reaction forces!

So you can see starting with vertical misalignment, this model predicts horizontal forces. Probably the predicted static forces could easily be resisted based on static coefficient of friction calculation but then there's the dynamic forces which tend to encourage relative movement also (how do you break free parts that are stuck together?... wiggle them).

The story I'm trying to weave is that the higher hold-down bolt torque may be required to resist horizontal movement based on known thermally-induced vertical misalignment expected in this application. It's just a story though… could be way off base.


=====================================
(2B)+(2B)' ?

 

[desertfox]

hi electricpete

Thanks for the response I'm looking further into it now.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[BrianE22]

Of course, if you prevent movement from thermal growth then you will have higher locked in forces and resultant stresses. For axial loads I'd look at rigidly connecting the two bolts closest to the coupling and coming up with some sort of "sliding" joint on the rear bolts.

 

[Tmoose]

Hi desertfox,

Can you post a picture or drawing of the installation, including skid dimensions and close up the details of the motor support provided by the skid??

I'm thinking the thermal expansion they are trying to resist probably refers to axial expansion, not vertical.

=========

I'm kind of working on a project now where a big name fan company seems to be relying on under-torquing the NDE of a hot fan housing to allow thermal growth in a controlled fashion. The fan bearings and motor are bolted to a pedestals welded to a frame that resembles a slice out of a 1928 Packard chassis.

http://archive.lib.msu.edu/MMM/EB/01/e/EB01e802.jpg

Two l-o-n-g channels with nary an X member or any other really torsionally stiff details. The whole shebang is (wishfully) Mounted on a 20 mule team of carefully shimmed isolators.

So far the outboard sleeve bearing has shown signs of 1X (shaft to bearing via prox probes) and very low Frequency instability (housing via accelerometers) sufficient to trip the machine's controls when things heat up, and the NDE fan housing has slid several mm sideways relative to the bearing pedestal to which it is mounted.

Some basic analysis suggests that, even with positively dainty torque applied to the housing bolts, before the fan housing mounts would slip, the NDE bearing pedestal would move axially about 1/4 inch and tip the bearing bore several thousandths out of line.

I agree with others who have said if the mounting feet must slip, they must be carefully guided, ala the keyed joint in hot Boiler feed pumps etc. Lest they move sideways as easily as purely axially.

thanks,

Dan T

 

[desertfox]

Hi Tmoose

Thanks for the response, sorry I am not in a position to post any pictures at this point in time as the drawings are unfinished.

However you have touched on axial expansion which is what I'm interested in, now it appears to me there is no intention to allow the motor feet to slide, the motor weighs 62 tons and in addition to M64 bolts which hold the motor down they are having us drive a taper pin through each foot into the fixed base.
I always thought that the axial expansion was taken care of via the clearence sin the gear couplings and the gap between the shaft ends.

I been recently informed that the bolts are to be tightened to 90% of the yield stress which again suggests they don't intend the feet to move, unless you know something I don't, strangely enough some of the engineers at the motor manufacturers agree with me that we don't need the M64 bolts and that the M56 bolts would be okay, however it's not their decision

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[electricpete]

You mentioned taper pins... are they at only the coupling end of the motor or are they at both ends?

62 tons = 124,000 lbm... wow.
Just curious, what is the speed and horsepower of this beast?

=====================================
(2B)+(2B)' ?

 

[desertfox]

Hi electricpete

The speed is 185rpm and the power is 5600kw.
The taper pins are one at each foot pad of the motor.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[Tmoose]

Hi Desertfox,

What is this motor driving?
Has the motor supplier ever built anything like this before?
Is the motor rotor horizontal, something like this?

http://www.corrosionfluid.com/assets/images/total_pump_solutions_header.jpg

In OP you said this.
"we were informed that the bolts need to be increased in size to M64 because of thermal expansion of the motor casing."
I think written elaboration with sketches and calculated thermal expansion from the supplier is necessary to understand what the heck is meant.

=================

"I always thought that the axial expansion was taken care of via the clearance in the gear couplings and the gap between the shaft ends."
For a motor with ball bearings this is mostly true since one of the bearings will be clamped in a position determined by manufacturing to serve as a locating bearing to position the rotor close to mag center. (The other ball bearing is expected/hoped to slide and painlessly accommodate whatever the differential expansion may be between the rotor and frame.) Bi directional Clearance between the motor/driven machine shaft ends assures the thrust bearings will not be subjected to the unrelenting and viciously powerful axial expansion of steel shafts. No thrust bearings are stout enough to handle that.

Right or wrong I imagine this large motor uses sleeve bearings. In the more modest sized sleeve bearing motors I have dealt with there is no thrust bearing operating full time in the motor. The job of axially positioning the motor rotor in operation close to magnetic center falls on the use of a limited end float style coupling and the thrust bearing in the driven machine. Setting the position of the motor housing and possibly varying spacers between the motor/driven shaft ends are required to end up with the motor rotor close to mag center

 

[PudumaterX]

I can just imagine holding the feet of the motor while the casing expands axially and plumps out and possibly even bending the casing causing an eccentric rotor condition... etc.

I wait eagerly to get the final solution.

 

[Euripodes]

I read some not all replies. I think the question has been confused. You must look at the total system, total forces, total mounting detail. Your locating pins are essential in the scheme of it all. This, below,is about the mounting system

The turning forces on the motor vary continually with pump load, voltage changes from energy authority, heat and friction forces. The motor transmits that torque all the way to the base and plinth.

It may well be that there are motion, sound absorbent devices disconnection-facilitating even splined couplings along the shaft and that some of the resulting forced are turned into heat and strain at that point. That's not our issue other than their reactiveness. which will automatically be covered if measurements are done of all the engineering outcomes during running..

Of course the motor and shaft and the other components heat up and for two or three reasons and they then could cause expansions acting against the mounting bolts on top of the normal torques.

Down at the mounting bolts: They have to match the torque forces...preferably not excessive in tension as they might then themselves snap or break the motor casting........... or too loose causing perhaps turning motion on the bolts, loosening them over some period (sometimes the arrangement is pinned or uses castellated nuts)

Heat can also aid in loosening bolts through the incremental expansion and contraction

Bolts stretch...that's how the required system tension is generated...against what should be mounting points of 'fixed elasticity'.

Putting that together...it's possible the wrong bolt size or tension was in the initial mounting and the company used the excuse to just fix that rather than tell the truth and face uncertain human nature.

It's also possible that to properly react to the pump forces at the bolts,which cause stretch-forces against the humanly applied tension ...(always I hope done accurately with a certified tension wrench in the correct bolt order)....have been deemed inadequate...

It could have even slipped over into the next higher tension 'bracket' owing to due diligence reviewed 'in operation' indicators...which might include a dial gauge indicating machinery movement over a range of real, as opposed to theoretical criteria, vibration measuring and periodic motor tension tests to indicate bearing condition.

Bear in mind that every time the torque exceeds the opposing tension of the bolt, the bolt will stretch and/or the motor mount 'dent' and reduce thickness.It could be that under bolt-head and nut that paint, plating or rough casting might fragment, even microscopically, reducing bolt tension

When going to a larger diameter bolt you should check with the motor manufacturer as to the reduced mounting size because if the original bolt was actually smaller than the mounting hole and the hole did not have to be enlarged for the mew bolts then that was a faulty installation at the outset. You should match hole sizes precisely, not 'minimum bolt size'

Drilling or reaming the hole to be larger in diameter might weaken the base or place the bolt head on uneven surfaces...the bolt head and the nut should be against perfectly flat and perfectly parallel surfaces...you must check any modifications with the motor manufacturer. If the tension applied surfaces are not perfectly flat and parallel some people would increase the tension, leading to perhaps, an eventual collapse at the mounting.

Some large motors I installed as an apprentice used bolt systems which incorporated springs to maintain tension and allow 'independent suspension' to a degree. I can't over emphasis the importance of correct torque Vs tension readings, use of certified tension wrench and absolute parallel of all mounting surfaces, bolt head and nut.

Never use spring washers...very few know how to use them...they are not meant to be flattened or to be used in accurate tensioning systems. The Belleville washer is the correct item to use for electrical connections say at switchboards but it must be known whether the reduced contact area of the Belleville washer will affect the required area/tension.

Belleville washers are not meant to be tensioned flat.

That leaves mostly the use of scientifically evaluated, matched springs under the bolt heads, tension wrench tensioning and a nut-pinning system or simply tensioned nut/bolt arrangement. Loc-tite or equivalent...ask manufacturer's warranty in that precise situation or ask rep to come and give a decision. The problem with castellated nut is not only possible weakness but that few H/T bolts have more than one hole, it has to be correct for the assembly, that split-pins can shear, and that to line-up the tension might be upset.

Using flat washers can be poor engineering as they may crack or reduce size. Everything has to be correctly engineered for best practice. Continual improvement (or PPI)requires monitoring and review each week for a months then as things settle....perhaps every month to 3 months and then six monthly. I hope this explains all the matters pertinent to you dilemma. Anthony Clancy,Eng.

 

[desertfox]

Hi Tmoose,Euripodes

The motor will be driving into a gearbox which has two output shaft connected to two steel rolls and yes the load will vary but the torque cycle would have been given to the motor supplier at the outset.

The bolts have been calculated based on the motor having an internal short circuit which produces a force on each side of the motor of something like ten or eleven times the force under normal operation, I don't have the figures to hand but I know the M56 bolts were capable of handling this, we haven't received a great deal of information from the motor supplier as yet, despite my numerous emails trying to ascertain more information, the latest info received ststes the bolt grade of 8.8 and a preload of 1400kN that's about 140 metric tonnes, the motor itself has a mass of 62 Tonnes so each of the four bolts will have a tension of twice this mass and then it's pinned in position by four taper pins, it doesn't appear to me that the base of this motor is being allowed to float under any kind of thermal expansion, in which case I cannot see why the M56 bolts are of no use.
To put this in a nutshell we have a motor held down with four bolts and someone is basically saying that the difference of the installation being wrong or right is the difference between M56 and M64 bolts,I'm struggling to see that.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[Euripodes]

Hi desert fox, you might have skimmed my post... I've answered everything. The most important thing of all is that your agent(?) wanted to uprate the bolts and wouldn't be doing it (free?) because he was bored on his day off. The inertia of the motor on start would be massive....I don't know whether it is pump on line and what sort of pump (has high load on start)and whether you have 'quadratic' or some soft start controller of course.

What I said about bolt holes (etc) is seriously important If the holes were bigger than the bolts by 9mm I'd be changing them too....irrespective of the pins...It may not be that the M56 is 'no use' but rather that for a reason you can ask their head honcho they decided on a change. I'm not sure of course why you are concerned about it...it seems they didn't charge you. It's not so much an issue of total bolt tension but each bolt tension as the motor has a variety of turning moments and each bolt feels the stretch force.

As far as what went through someone's mind in making the decision that's outside my line of work. Say to them 'your reasoning doesn't make sense to me...would you please give me a full math's reason for what you did. For me a wrong hole size would be enough with a machine that size...but we could get a 'Eureka' and explain exactly a sufficient reason to do it and still be a country-mile away from reason that actually drove it. Maybe they wanted '100%' safety or '200%' (there is no such thing as 'over 100%' in % terms but anyway....)

On one of my jobs an arrogant Croat expert borrowed 4 of my blokes (unbeknown to me) in the tunnel at GID he had a 40 ton winch to pull some cables. By the time he'd finished it had 6 inch x 3 inch bolts set into concrete and every safety and overload over-ridden. The winch pulled out of the concrete and it went through the air like a bolas...hitting one of my guys in the chest as it went past him. The forces can be massive. I won't contribute more on this but the answer will be somewhere in my contribution. and others have probably covered any remaining logical possibilities To get their reasoning.....ask them directly or for a written account. You know...there may not actually have been a legitimate reason,..... but whether or not, M64 isn't a huge bolt...did you have to enlarge the hole?...is M64 the correct close fit size?

If you want to send me the provider details, I am quite happy to aske them myself.....if not....remember that one day you'll be able to ask St Peter....(LOL!!) Good luck with it all....cheers chaps.

 

[desertfox]

hi Euripodes

I did read your post but its nothing to do with the bolt hole size, the motor drives through a gearbox which connects to two rolls as I mentioned in my last post.
I've asked for information to back up their logic but none as been forth coming, what I originally asked for was any information on thermal movement on the motor feet relating to bolt size because I could not find anything.
sorry I can't see how you've answered everything but maybe ST Peter will know!

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[BrianE22]

Europodes -

There are two reasons for the bolts - 1) to handle the torsion of the motor and 2) to create enough pre-load (and resultant friction) that the motor does not move axially under thermal loads.