Procedure For Alignment

Procedure does not mean much without some idea of the tools available to you...

Do you have a laser alignment system? Then follow the instructions in the computer

Do you have a reverse dial indicator rig? If so there is a fairly rigerous procedure to be followed that we can discuss if you wish.

Do you simply have a standard dial indicator? Then a rim and face approach is probably appropriate.

If the only thing you have is a straight edge and a level, you can get close, but not very close....

So again, what tools do you have?

Also, the procedure of a turbine alternator (flange style rigid coupling) is very different than a motor driven pump (flexible coupling). What kind of machine are you really interested in?

I have not personally been involved with an alignment, but some of the things that come to mind are; The bearings have to be aligned for height, this done with a water tube, like say a turbine and alternator, from memory I believe a small dish is placed on each bearing housing , and each being connected , water is slowly added till there is an equal amount in each dish, someone with experience will need to elaborate on that. For overall alignment a piano wire is tensioned from the first bearing to the last bearing, and bearing housings adjusted. If white metal bearings the shaft or rotor will need to be blued, and the bearing scraped in. Multiple coupling will need to be aligned with clock gauges and feeler gauges. This is just a rough outline hope it helps.

Regards JD.

we bow to your wisdom

Not until I'm a guru.

Regards JD.

Step by step procedure is explained in our website: www.yesyen.com > Downloads > free downloads > graphAline.

I don't quite know why, but I just love this post of JD as one of my all time favorites !

You know why you love jd's post

nudge nudge wink wink

Your reply was my second favorite, and his reply to yours my third. At least we have 2-3 real gentlemen on here. I like it.

It also depends on scale. We work with 20 mm diameter motors and a plastic disc coupling to a ball screw. We use comparators on a surface table to line all companents up to within 20 microns that way. If it is out we use shims but if the diference is too small, we use celotape which you can compress to the right thickness.

In larger applications we just clocked from frame to motor mount or from shaft to shaft. That way you can get it close enough for the rubber couplings to take up the difference. I expect it to be somewhat different in turbines as the rpm is a bit higher. Here you need to be within microns again but over much larger distances.

Alignment concepts are simple. The tools available only help your shafts arrive to your understanding of what alignment really is.

I teach the following in my alignment class:

1) Know where your shafts are when you are performing the alignment

2) Know where your shafts will be while running at the expected load

3) Close this gap so they are in alignment while running keeping in mind you aligned them while not running

***********

Equipment cases (pumps, motors, gearboxes, turbines, engines, etc.) normally hold the shafts in suspension. Therefore if those cases move, the shafts will move with them proportionally. So if you are performing your alignment with tools mounted to a shaft held by a cold case, and while running that same case becomes very hot, the case will grow and the shaft will be out of alignment while hot. Therefore you must intentionally misalign the shafts cold, so that they are in alignment hot. This is termed 'thermal growth'. It can be estimated for different materials. Normal steel grows at a ratio of 0.0000065 inches per each degree Fahrenheit differential. Search Google for "coefficient of thermal expansion for X" where X is your case material.

Keep in mind your alignment is performed mostly at ambient temperatures with machines at that temperature. So an electric motor will get very hot and grow, while a pump that is pumping very cold water may shrink when pumping.

Next, you must consider where your shaft will be dynamically (running) inside the bearings. Shafts running in sleeve (journal) bearings ride on top of an oil film, so, they move while running just due to running on this film. Shafts move due to dynamic loading also. For example a horizontal split case pump shaft may move up and towards the discharge flange due to hydraulic loading of the rotor. But, ball bearings are tight when hot so they hold the shaft very centered, however, a large ball bearing (8" diameter for example) when cold will allow the shaft to rest at the bottom of the bearing by 0.001" (one third the thickness of an average human hair).

For small equipment such as 500 kW the sleeve bearing dynamic shaft movement may be insignificant but for 15 MW 12 " shaft for a sleeve bearing gearbox, the shaft may move up and away to one side from dead bottom due to tangential gear separation forces, and this movement can be 0.008" or more.

So, know where your shafts are dynamic and hot compared to where they are at rest (when you are doing your alignment).

Some cases do not grow uniformly. Ex: A large diesel engine that is 5 m long will likely grow up more on non drive end and move off center to one side in the front also.

The actual procedure, using lasers is:

A) Calculate thermal position based on ambient temperature at time of alignment compared to running temp and using the thermal growth formula determine the approximate shaft positions hot, and use these targets to plot an intentional misalignment cold that will correct itself when hot.

B) Added to (A) above, overlay the dynamic shaft position given to you by the equipment manufacturer to arrive at an estimated position hot and dynamic.

C) Practice installing the lasers in the correct place (nearest to the bearings attached on the shaft with the coupling installed). Perform a cold alignment using the shaft target obtained in (B). Take off the lasers but lay them out near the mounting position because we are going to put them back on later very fast before the machine cools off (get ready to install them later in less than 5 minutes time).

D) Run the machine under low load, checking for vibration, and if OK go slowly to full load. If vibration is high, but not so high to damage things, allow the unit to warm up to full temperature and let it "heat soak" for 2 hours after temperatures have stabilised.

E) Shut down and quickly put on the lasers and check alignment hot with gathering your new 'hot' shaft position data within 10 minutes (so the unit doesn't cool down and contract).

F) Compare this new reading to the dynamic shaft offset information given to you by the manufacturer (B above). i.e. if you know you will have dynamic shaft offset, and the unit is hot, and the shafts are perfectly aligned at rest, you are OUT of alignment.

The hot alignment removed the thermal growth estimate and replaces it will real life 'hot' situation. The thermal growth calculation is only an estimate and these numbers can be off by 50%. So a hot alignment removes those math errors.

Further alignment concepts such as surface roughness of mounting feet, dowel pinning, shim crush, dynamic case deformation measurements, shim quantity, surface finish and hardness of shims, grout, liquid shims, etc are beyond a simple post and would cause my wife to unplug this computer.

And just to know, most alignment books are so old they are not very helpful (Steve S. wrote most of them when he was a cabin boy for Christopher Columbus). Search for "laser shaft alignment" and manufacturers of such equipment will appear and training companies will appear and there you will find lots of info.

Take care

.........nice post ...and ...tutorial...thanks for the information , i have not done that huge alignments , just smaller machines say sleve coupling in water pumps ,engine generator with clutch coupling , they are obselte now with monoblocks and factory set alingnments......

Hey! I will tell you this... If you are looking to align a pump/motor, turbine, turbine/compressor train, generator, etc., you need to understand first off that precise alignment is of the utmost importance. In a plant environment, the products that are created/moved by these machines are usually not just water or air. It is not something you just decide you can do one day. Someone mentioned straight edge alignment. Unless you are talking very low rpm's, and a machine with a flex coupling, the straight edge method should only be considered the first step of the process. For most rotating equipment, especially high speed, you are looking for tolerances of no more than .001, though some companies will buy off at .002 or .003. Personally, I feel that if you can get to .006, you can get to .003. If you can get to .003, then you can get to .000. It just depends on your patience. The more precise, the longer and better it will run, (barring other issues) as the longer your bearings will last.

Two dial indicators with mag bases or chain clamps work fine with the graph method, but are difficult for a novice to understand (double reverse dial, or rim and face). If you have access to Optilign, Rotolign or Rotolign Pro, and are somewhat computer literate, it will be much easier for you. These machines, as long as the setup is correct, tell you exactly what moves to make, shim changes, if you have a soft foot, etc. I'm sure they can be rented, though most plants have one on hand if they employ millwrights, or do their own millwright work. Don't forget lock out tag and try before you open anything. In my opinion, this should be done by a professional who is proficient at it. I've seen disasters from cutting that corner. Good luck.

"tolerances of no more than .001, though some companies will buy off at .002 or .003. Personally, I feel that if you can get to .006, you can get to .003. If you can get to .003, then you can get to .000. It just depends on your patience"

This statement is only true if you fully accept that if you read .003 it has to be corrected with the device's deviation. This can be considerable depending on the system in which you apply it and the skill of the person handling it. I assume these tolerances are in inches.

We work on 0.03mm machining tolerances and then have to clock to 0.01mm the allignment of ballscrew in 2 planes. We use mitutoyo comparetors with a small ball nose which can read out to 0.005mm. Often when the product fails repeatability tests, we find that even though you read the dial right, the real measurement was out due to stacked up errors. No way around those as you have to include a percentage fall out on expected errors.

Don't just read the device without taking its limits into consideration. You will only fool yourself into thinking that what you thought was achieved is real. Big mistake.

I stand by my previous statements. Period.

Whereas I agree in principal that close alignment is better than rough alignment (0.000" v 0.006") but in the real world we justify increased mean time between failure compared to lost revenue while performing a precise alignment.

Let me illustrate. To align a 500 ton machine to 0.000" is extremely difficult to do and may require 3 more moves at 2 days per move. If this machine was a critical pump on a one million per day pipeline, the cost of going from an acceptable 0.003" to 0.000" may be $400,000,000 in lost revenue and the investments that can be made on that revenue in those few days as well. Now will a 0.003" run $400,000,000 better? No. Both the 0.003" and 0.000" will run an equal time between failures. Will a 0.006 run as long as a 0.003"? It depends on the machine rotors and what they are used for. It depends on the bearing clearance and how the rotor movement affects other things in the train. A gear box will not like this much offset as the bending stress goes up exponentially at the gear teeth with only 2-3 mils (0.001" = one mil in USA slang) difference. But, as long as that stress is well under the limits, who cares! But the gear box case vibes might go up due to gear teeth meshing out of alignment and those vibes might excite a pipe that breaks and causes a fire.

However, to those that say close alignment is not required on a 500 ton unit I am quick to point out that the overall professionalism of the site is raised by performing a close alignment with PROPER tools. Machinery professionalism = longer life. Simple. But moreover, I've seen the fascination with close alignment tolerance become covered by lack of knowledge of basic alignment concepts; that being a true and thorough understanding of where the shafts are while running under load. A precise 0.000" alignment with rim and face clocks might make us feel really good that we are 0.000", but the shafts may be off while running by 0.020" if the boys don't understand that rim and face is VERY inaccurate for most complex machine trains and if no intentional mis-alignment is made with shafts at rest then we have a concept problem (although in our head we think we are 0.000).

So what is proper alignment? It is alignment that produces acceptable vibration limits, case deformation and / or shaft positions while running. To get tighter than that is an insurance policy and must be evaluated on a cost / benefit basis.

I totally agree with what you're saying, for the most part. It depends on what machine you are talking about. There is no way in hell, for instance, that a customer (say, BASF) would accept .004 on a steam turbine, centrifugal compressor train. And they shouldn't. If you're working on it, there's a reason it is down in the first place. The idea is to only have it go down for shutdown (maintenance). Whatever it takes to get it right while it's shut down, is what is accepted, usually. And yes, when I say .000, I mean that lift, thermal growth, etc. have been calculated into the final. I've been on a turnaround where they didn't want to spend time getting a good alignment, just to go back 2 days into startup because it kept throwing the breaker due to vibration, and after 3 tries, you know what happens. Anyway...

There are multitudes of machines out there that must be aligned properly, and in no way have I worked on all of them. But I stand by my statements made previously.

And that is some pump to be 500 T. I'd love to see pics of that.

I agree with your thoughts also. But faced with 3 more moves at 2 days each to get another mil or 2 when I'm way inside the ballpark I'll take the chance and monitor shaft orbit. The 500 ton units would be extreme compressors. In my world of large engine driven pipeline pumps (my niche), this is a typical driver:

http://www.wartsila.com/Wartsila/global/docs/en/power/media_publications/brochures/engines/wartsila_34SG_engine_technology.pdf

They have larger engines. They removed pics of 2 of my projects from their website (showing the entire machine train) because those engines are replaced by current models. When you have a fixed pump, a 1:6 gear (6 bolts at 60 mm each) and 16 engine bolts at 42 mm each, it is real depressing when you move the gear to clean up the alignment by .002" better and then must move the engine to match, then during the re-run for hot check you find you need one more move of the gear (and thus engine).

I'm a pipeliner and upstream guy. I'm just now learning downstream and the culture of plant managers and I must say I not too impressed. The sites regularly do things I'd fire people for in my world. Totally different culture (and a different profitability business model).

Take care

Very impressive. I've never seen one, as I am in the refinery world. I think the largest rotor I've seen was 104 tons. (gas turbine, modified). I know what you mean, though. Smallest, maybe a grand. 5 1/2" nuts on case bolts sure make me thankful for bolt stretchers. 70 or 80 of those, whew. Can you imagine hammer wrenching those? Even hytorq kicks your butt. Thanks for the pics. Couln't see the specs, but again, very impressive.

For case closure nuts I'm now recommending Riverhawk hydraulic nuts which are bolt stretchers built into the nut (a pancake jack built in). It's like a hydraulic super nut. Screw on 50 or so, attach 50 hoses and pump them all up at once evenly then walk around and spin down the retaining nuts and release the pressure. Shazam.

Open this and look on page 11. These big engines do everything with hydraulic stretchers. Rod nuts, main nuts, head nuts, all on assemblies that stretch all 4 or 6 head studs at once with one hydraulic pump, etc., etc. The main bearing is fed oil from a rail below each main cap down the middle of the oil pan and individual hydraulic jacks attach to each main cap from this rail and lube oil is pumped up through the center into the main bearing and through the crankshaft (reverse of normal). Then when changing main bearings, you attach your stretcher to the main cap studs, spin off the nuts by hand, and hydraulically lover the main cap down on this little jack and pop in a new shell.

Would be a cost prohibitive change for most refineries, when the same can be accomplished for and F7 gas turbine or large centrifugal compressor (so many nuts involved) using the bolt stretchers they currently use. I don't see them changing, when it can already be done in one shift. Would be nice, though.

When I said disasters, I mean dead people.

are you doing reverse alignments?

Maybe this experience of mine is related with alignment or not.

I erected a Tobacco Rotary Drier at Pakistan Tobacco Comapny ( a B.A.T Industry) 9 ft dia x 23 ft length, aligned it, levelled it. When the englishman erector of BAT(British American Tobacco) came he had to show he was superior because their shit dosen't smell. He started the Rotary Drier and asked for a glass of water (which I thought to drink) but he put the glass on the Gear Box. The water in the glass was vibrating and he said, "No good - Vibration re-erect machine again". My Technical guy came running for me saying, we have to re-erect the machine again. I went there, took the glass of water off the gear box and placed it on the ground. The water in the glass was vibrating in the same propotion as on the gear box. I told the bright englishman, " You stop this and I stop that". He stood there scratching his head and left for the guest house. In the afternoon, I heard he left for England. Haven't heard form him since then.

No offense except to smart asses only.

The Machine is been running since last 9 years with no problem till date.

Just though I'd share this personal experience so that all alignment may not be wrong, the floor vibrates with the weight & rotation which we tend to forget and take into account.

Caution: Remove all liquids from your mouth before reading unless you have a Panasonic Tough Book.

My most favorite of all is the famous coin test !!!!!!!!! Gosh . . . I love the coin test.

Stand a coin on end on the machine case and if it doesn't fall over all is good right ?

I just can't wait for inspectors to pull out their coin!

I took some chewing gum and placed a small amount on my coin edge and hung my coin upside down UNDER a bearing housing.

That THAT baby is aligned PERFECTLY !!!!!!

Rollin! I don't know where that test got it's start, but I've seen it used a few times.

I took some chewing gum and placed a small amount on my coin edge and hung my coin upside down UNDER a bearing housing.

That THAT baby is aligned PERFECTLY !!!!!! - You kidding

Hi, ducon!

I'd really like to try this method, as it sounds simple and accurate. What kind of gum is best for the application?

Mark

You flatter me gentlemen, I must admit my first alignment was of a triple explanation marine steam engine shaft,renewing the white metal bearings, and that was before lasers, if you wont to know how alignment was achieved, it was done with a piano wire and strips of triangular metal, that where fitted under the wire without disturbing the others, the bearings where machined accordingly and scraped in. A long long time ago. I see the essence has been added by Petripower, very nice.

Regards JD.

I've seen this done in a paper mill to align the rolls. I loved watching it. Don't know how much they do this anymore, though. We also don't usually scrape babbit bearings, but laby seals are a different story.

Hi, jdretired!

That was back in the days when the first thing you learned to do was make all the tools you needed to do the job! (Unless you had some put away from the last time!)

Mark

There are no real standard procedures for alignment as per alignment of rotational equipment. This is due to factors of thermal growth, pump/ compressor/ turbine design/applications, what equipment is being used to drive these systems and how they are connected. In most cases each job has it's own characteristics that are not similar to the previous job.

Factors such as ambient temperatures, indoor/outdoor situations, sun light influence, vibration issues, base issues, all of these issues arefactors of the job.

You can use an outline document for basic procedures for this type of alignment and then modify to the application you are working on.I hope this helps a little.

As far alignment equipment - dual dial indicator method, computerized laser alignment, laser tracking systems, various optical systems can be utilized.