Handbook Of Shaft Alignment Episode 1 Part 6 pot

Once theentire baseplate was filled, the grout began to flow and seek its own level and we noticed thatthe grout began to swell higher in the pour holes at the low end of the baseplate a

The four bottles of hardener arrived and the pour began in earnest at 1245 h Theprocedure was to blend the liquids (epoxy and hardener) for 3 min and then slowly addeach bag of aggregate to the mixture as shown in Figure 3.47 Once the fourth bag ofaggregate was added, another 2 min of mixing was suggested before the grout should bepoured An electric drill with a mixing blade was used to mix the contents When the firstbatch was mixed together, it became apparent that the mixture was very viscous (almost likepeanut butter) The drill motor quickly became overloaded and the windings began tooverheat and smoke Another drill motor was at hand so the drills were swapped out and

by the second batch, it too became overheated A larger drill motor was quickly found tohandle the mixing with one batch mixed at a time After adding all the contents, the barrelweighed around 250 lb, a little too heavy for two people to lift and pour directly into thebaseplate

Because of the viscosity of the grout, the most effective way to pour was by hand, scoopingthe grout out of the mixing barrel and then pouring it into a hole as shown in Figure 3.48.After the barrel got half empty, it was light enough to be lifted by two people We then wouldfill up a 5 gal bucket and use it to pour while another one or two people would continue toscoop it out with the smaller buckets (which were made with thick plastic 1 gal laundrydetergent containers) Thankfully, two barrels were available where one crew would pour thegrout whereas another crew would mix the next batch Initially, each batch took 15 min fromthe time the mix was started to the time the barrel was empty After a quick calculation, it wasFIGURE 3.46 Protecting the baseplate and surrounding area with plastic covering

FIGURE 3.47 Mixing the grout contents

conveyed to the work crew that it would take 4 h to finish the pour, far in excess of the curetime so additional personnel were called in to speed up the process as shown in Figure 3.49.About the time the fifth batch was poured, we noticed that the wooden form around thebaseplate started to leak as shown in Figure 3.50 Thankfully, we had the foresight to havesome duct sealant at hand in the event that this happened and the leaks were plugged asshown in Figure 3.51 It then took approximately 6 min for each batch with bodies scramblingaround feverishly mixing and pouring the grout At 1530 h the final batch pour was made asshown in Figure 3.52.

The top plate of the baseplate was designed with a slope toward the pump side so in theevent of a water leak, the water could drain off the base to prevent rust from damagingthe baseplate Knowing this, we started the pour at the low end of the baseplate Once theentire baseplate was filled, the grout began to flow and seek its own level and we noticed thatthe grout began to swell higher in the pour holes at the low end of the baseplate and dropdown at the high end as shown in Figure 3.53 As seen in Figure 3.54, we then topped off thepour holes at the high end hoping that the epoxy would begin to harden at the low end wherethe pour first began to prevent it from overflowing onto the top of the baseplate The groutindeed did begin to slowly harden and, in fact, became somewhat like a soft putty so itwas decided to begin carving off the grout that had leaked out around the form as shown inFigure 3.55 and Figure 3.56

FIGURE 3.48 Pouring the grout

FIGURE 3.49 More people needed

FIGURE 3.50 Grout leak around form.

FIGURE 3.51 Plugging the leak with duct sealant

FIGURE 3.52 Finishing the pour

FIGURE 3.53 Grout began swelling at low end.

FIGURE 3.54 Hand packing the grout

FIGURE 3.55 Scraping off the excess grout while still in the putty stage

The majority of the work crew then cleaned up and left Two of us decided to stick around

to smooth off the grout at the pour holes as the grout began to harden, which began toquickly accelerate about 1 h after the last pour had been made Since the chemical reaction ofthe epoxy grout is exothermic, the baseplate began to get warm, then somewhat hot We alsobegan to notice that the epoxy began swelling out of all of the pour holes apparently fromexpansion of the grout during hardening Surface temperatures of the baseplate were takenwith an infrared pyrometer with temperatures ranging from 1278F to 1398F The epoxy began

to harden very quickly around 1700 h, so it was decided to remove the protective plasticsheeting from the top of the baseplate before the epoxy hardened completely The epoxy hadalso oozed out of the vent holes and by this time we had to chisel them off as shown inFigure 3.57 Figure 3.58 and Figure 3.59 show the grout pour holes after the epoxy had cured

A fan was placed to begin cooling off the baseplate overnight

The next morning it was decided to take another set of optical alignment measurements onthe 8-ft pads to see if the baseplate had stayed in the same position prior to the addition of thegrout Figure 3.60 shows the jig transit and optical scale target on the foot pads The transitwas precision leveled and the line of sight was adjusted to buck back into the same elevationplane by observing the adhesive backed target placed on a nearby building column when thefirst set of measurements were taken as shown in Figure 3.61

FIGURE 3.56 Scraping off the excess grout while still in the putty stage

FIGURE 3.57 Chiseling off the excess grout

FIGURE 3.58 Epoxy at grout pour hole.

FIGURE 3.59 Baseplate after clean up

FIGURE 3.60 Jig transit set up to observe final elevations on foot pads

Figure 3.62 shows the elevation data and the baseplate profiles before and after the grouthad hardened Figure 3.63 shows the pump to be installed and Figure 3.64 the turbine to beinstalled onto the baseplate.

As shown in Figure 3.62, the baseplate had distorted after the grout had been poured.Notice that pad E did not change its position very much after the pour had been made All ofthe other pads changed their position with the pads in the center of the baseplate now muchlower than either end The baseplate bowed downwards more in the center, a little at the westend, and virtually none at the east end

The following conclusions can be made:

1 The top surfaces of the four pump foot pads were, and still are not in the same plane

2 The top surfaces of the four turbine foot pads were, and still are not in the same plane

3 An 18 in long precision machinists level is unable to span across two of the pump orturbine foot pads to check for longitudinal and transverse levelness

4 If a precision machinists level would have been used, the leveling process would havegone on forever Depending on which pad the machinists level was placed on and whatdirection it was placed in, the baseplate would have to be re-leveled for each pad Sincethe pads are sloped differently, once one pad was precisely leveled, when the machinistslevel was moved to another pad, it would be out of level

5 Every effort was made to position the pump foot pads and the turbine foot pads in anaveraged level, coplanar, and parallel condition prior to grouting This was achieved onseven out of the eight foot pads This could not have been achieved by the levelingjackscrews alone In several cases, the anchor bolt nuts had to be tightened to bend thebaseplate downward to achieve the desired elevation at certain foot pads There were nojackscrews or anchor bolts located at pad D to distort the baseplate at that position Allanchor bolt nuts were tightened and the adjacent jackscrews tightened to hold thebaseplate in its pregrouted position

6 Every effort was made to follow the installation guidelines set forth by the equipmentmanufacturers, the grout manufacturer, and the procedures set by API RecommendedPractice 686 The objective was to completely fill the baseplate so the grout would bond

to the top of the concrete foundation and the underside of the baseplate

By carefully studying the before and after baseplate profiles in Figure 3.62, it becomesobvious that the baseplate changed its shape after the epoxy grout had cured Our initialFIGURE 3.61 Reference target on building column used to buck in to same elevation

thoughts were that the baseplate may have moved upward due to the temperature increasefrom the exothermic reaction of the epoxy Instead, the opposite happened and it was not alinear move Upon curing, epoxy grout shrinks Once the baseplate was filled and the epoxy

10 in.

Final pre and grout surface elevation and profile

Surface with north edge higher

Pump

Steam turbine

17.805 in 17.799 in.

17.808 in 17.814 in.

No 15 BFW baseplate before grouting

Pregrout “shoot for” elevation

Post-grout best fit curve

Pad D Pad B

bonded to the underside of the top surface of the baseplate, the grout shrunk and bowed the1=2 in thick top surface plate downward in the middle despite the fact that there were severalstructural steel cross members in the baseplate design (see Figure 3.29).

Discussions took place on how to fix the out of level and noncoplanar surfaces of the footpads now that the baseplate was grouted Suggestions were forwarded to field machine all ofthe foot pads to make them level and coplanar If this was to be done, optical alignmentequipment should be available to assist in periodically measuring the surfaces that would bemachined to achieve level, parallel, and coplanar foot pad surfaces This however would be awaste of time and money Getting the foot pads flat and in the same plane assumes that thesurfaces on the underside of the pump and turbine are flat and in the same plane Is this true?

No data was taken to verify this despite the comments of one of the equipment manufacturers:

‘‘This couldn’t possibly happen.’’ If you look at the photograph of the turbine in Figure 3.64,you will notice that the turbine supports that will touch pads D and E are L-shaped platesthat are axially bolted to the lower turbine casting If these bolts are loosened in the casting, it

is possible that these support plates could be moved due to any clearance between the shank

of the bolts and the holes cut into the support plates

It was decided to set the turbine and pump onto the baseplate without machining and checkfor any soft foot conditions using the procedures described in Chapter 5 Figure 3.65 showsthe soft foot map when the turbine and pump were set onto the base Assuming the undersides

of the pump and turbine feet were flat and in the same plane, there should have been very littleFIGURE 3.63 Pump being installed on baseplate

FIGURE 3.64 Turbine being installed on baseplate

(if any) soft foot problems on the pump (pads A, B, F, G) Now look at the soft foot map inFigure 3.65 and observe that the gaps at pad A indicate that the pump foot was not makingcontact there Therefore the undersides of the pump feet were not in the same plane A similarstudy of the turbine foot pad profile (pads C, D, E, H) and the soft foot map will illustratethat the underside of the turbine feet was also not in the same plane.

I have had the opportunity (i.e., been allowed) to use optical alignment equipment a total offour times measuring the four corners of all the foot pads and have seen similar conditions onevery baseplate checked this way I also know that very few baseplate installations are donewith this rigorous of a measurement process and that carpenters levels, not machinists levels,are frequently used and that very few people verify that a baseplate is indeed in level after theinstallers say it was I am also not sure how often a baseplate was checked for levelness afterthe grout was poured

It should now become obvious that this may indeed occur very frequently and that footpads quite likely have a tilt and or twist condition and that the surfaces are not coplanar andthat when the machinery is placed onto the uneven, twisted, tilted foot pads, that a flat piece

of shim stock will not correct a complex, wedge-shaped gap that will occur It is difficult, butnot impossible to fix this Chapter 5 will discuss the procedure for doing this

Finally, what effect will a baseplate that is not in level within 2 mils=ft and 5 mils across theentire baseplate have on the successful operation of the machine? If a drive system has a 50mil slope across the entire baseplate, will the thrust bearings not be able to accept this slightaxial force from gravity? I do not prescribe installing baseplate with that radical a slope, butattempting to achieve the tolerances set forth by manufacturers and professional organizationseems to be unachievable in the real world Many may think that what was observedduring this particular installation does not occur very frequently, when in fact, it is pro-bably quite common Perhaps some rethinking needs to be done in soleplate and baseplateinstallation specifications

3.3 PROBLEMS TO LOOK FOR IN YOUR FOUNDATIONS AND BASEPLATES

A complete visual inspection should be made at least once a year of all rotating equipmentfoundations, baseplates, piping, etc Many of these problems are quite obvious as shown inFigure 3.66 through Figure 3.71

No 15 BFW soft foot after grouting

25 29 0 0

0 0 0 0

3 0 8 3 2

0 0 0

Support foot left loose

FIGURE 3.65 Soft foot map of pump and turbine

FIGURE 3.66 Fan frame mistakenly designed with no soleplates imbedded in concrete resulting in nocontact between underside of frame and top of concrete in the inertia block.

FIGURE 3.67 During a torque check on an anchor bolt, it was discovered that the anchor bolt hadsheared and an attempt was made to tack weld the anchor bolt so no one would notice

FIGURE 3.68 During a torque check on a steam turbine sway bar, the bolts threaded into the underside

of the outboard bearing were found to be loose Channel iron was used for the sway bar and the shaped washers would loosen after a short period of time from the vibration of the machine The boltswere tightened, the vibration on the outboard bearing would be acceptable for a period of time but asthe bolts began to loosen again, the vibration would steadily increase to unacceptable levels

wedge-3.3.1 PIPING, DUCTWORK,ANDCONDUITSTRAIN

Piping strain is a monumental problem in industry, difficult to detect on installed piping, andtime consuming to correct The widely held design philosophy seems to be that piping should

be loosely constrained so it can move and grow wherever it wants Many people are surprised

to learn that the vast majority of piping failures have occurred from cyclic fatigue, not fromtension, compression, or shear failures Most of the piping supports in existence were installed

by pipe fitters who were just supporting the pipe before all the connections were made Similarstrain problems can also occur on ductwork for fans and conduit for electric motors.Excessive piping, ductwork, or conduit forces can:

1 Distort the machine case upsetting internal clearances between moving and stationaryparts of the machine

2 Cause the machine case to shift its position over a short (or long) period of timedisrupting the alignment condition

3 Cause the hold down bolts to loosen or shear along with the shim packs and dowel pins

if used

FIGURE 3.69 Inspection of the grout under the soleplate revealed that the grout had begun todeteriorate Apparently this was observed prior to this and precut shim stock was used to attempt tocorrect the gap between the top of the grout and the underside of the soleplate, which is not recom-mended Also notice that the top of the anchor bolt is not fully engaged in the nut

FIGURE 3.70 A misalignment condition between the motor and water pump caused the packing to leak.Over time, the spraying water severely oxidized the pump casing and baseplate

Static piping forces that result from improper fits cannot be detected by simple visualinspection after the piping has been attached to a pump, compressor, or turbine Looking at aspring hanger and seeing that the spring is compressed does not indicate that the load iswithin acceptable limits Also, spring hangers can only support piping loads in one direction.What if there are other forces acting in directions other than that through the axis of thespring?

Even if expansion joints or flexible hose sections are included in the piping, these devicescan only accept forces in one or two planes of motion Installing flexible piping sections may

be just an excuse for someone to do a poor piping installation Furthermore, flex hose is moresusceptible to failure than rigid pipe

FIGURE 3.71 A poorly designed frame fabricated from plate steel vibrated excessively and eventuallythe angle iron support gussets broke loose from the concrete support holding the gearbox in positionputting the unit into a severe misalignment condition The bearings failed on both machines and ratherthan remove the concrete pedestal and provide a complete support for the gear and the motor, a pipeand jackscrew were installed, which eventually failed after a short period of time

Forces from the expansion or contraction of piping attached to rotating equipmentcarrying fluids whose temperatures are above or below the temperature of the pipe when nofluid is moving can be enormous and frequently cause drastic movement in the turbomachinery from excessive forces at the connection points The flanges and connections onpumps, compressors, fans, etc., were never meant to bear the weight or strain of piping andductwork They are fluid connection points The piping must have adequate support mech-anisms that bear the weight and strain of the piping in the vertical, lateral, and axialdirections A good piping design engineer should never view a pump or a compressor flange

as an anchor point for the piping

I know that this has never happened at your plant, but I have seen pipe fitters attach a 20 tchain fall around one end of a pipe and the other end around an I-beam, pull the pipe intoplace, install and tighten the flange bolts, and then remove the chain fall Some of the piping

in industry is so poorly installed that the pipe fitter has to stand out of the way when the pipesare disconnected from machinery for fear of getting hit by the pipe when it springs away fromthe connection

Figure 3.72 shows an adjustable piping anchor support If the piping misalignment is tooexcessive and you are not willing to fit the pipe properly, you may want to consider usingsupports similar to this at the suction and discharge pipes on the pumps and compressors toprevent the pipe from forcing the machinery out of alignment If the piping strain is excessiveand there are no supports to hold the poorly fit pipe in place, there is no guarantee that theequipment will stay aligned for long periods of time even if you do a great job aligning therotating machinery

Feet of support base are bolted firmly to ground

Axial adjustment screws and locknuts

Horizontal adjustment screws and locknuts

Vertical adjustment screws and locknuts

Pipe

FIGURE 3.72 Adjustable piping support

3.4 CHECKING FOR EXCESSIVE STATIC PIPING FORCES ON

ROTATING EQUIPMENT

Since a majority of rotating equipment is used to transfer liquids or gases, the connecting pipingwill undoubtedly have an effect on the machinery and could potentially be another source ofmachinery movement due to thermal expansion of the piping, reactionary forces from themovement of the liquid in the piping itself, static weight of the piping, or piping that has notbeen installed properly causing tension or compression at the piping–machine interface.The forces that cause machinery to move from improper installation of piping can bechecked by using dial indicators to monitor both the horizontal and vertical movement of themachine case By placing indicators at each corner of the machine element, loosening all thefoundation bolts, and observing the amount of movement shown on the indicators, anyundesirable forces acting on the machine can be determined If more than 2 mils of movement

is noticed, it may be possible to reposition the other elements in the drive train withoutmodifying the piping to eliminate this problem

This movement can also be checked with a shaft alignment bracket attached to one shaftwith dial indicators positioned at the twelve o’clock and three o’clock positions, on theadjacent shaft as shown in Figure 3.73 or it could be check using a magnetic base and dialindicator set up to observe motion of the machine case itself as the foot bolts are loosenedwith the piping attached as shown in Figure 3.74 Shaft or casing movement exceeding 2 mils

on any dial indicator is unacceptable after all the foot bolts have been loosened

3.5 VISUAL INSPECTION CHECKLIST

On an annual basis (at least), the following inspection should be made on all rotatingmachinery at your plant site:

1 Cracked concrete bases or support columns

2 Cracks propagating at concrete joints

3 Water seeping between baseplate and concrete foundation that could freeze and damagethe structure

4 Loose foundation bolts

5 Shim packs that worked loose

6 Rusty shims

FIGURE 3.73 Excessive piping stress test (Align the machinery and then attach brackets or clamps toone shaft and mount dial indicators in the vertical and horizontal position against the other shaft Setthe indicators at zero, loosen the foot bolts holding the piped machine in place, and monitor theindicators for any movement Ideally less than 2 mils (0.002 in.) of movement should occur.)