Machine frame distortion vs. foot related resonance
There is been a lot confusion around the term “soft foot”. Vibration analysts often understand some pieces of what can be described as soft foot but not be aware of all the phenomena that can be referred to by that term. The common question is “What frequencies will occur from a soft foot?” This question indicates an incomplete understanding of what problems soft foot can create and sometimes all that the term can refer to. This can be broken down into two main areas, machine frame distortion and foot related resonance. These require different methods detection and correction, and sometimes they may both occur on the same machine.
The term “soft foot” itself is inadequate as it does not describe the situation. If taken literally, a soft foot would imply that the foot is not hard enough. Of course, most people have a better understanding of what is meant, that the machine’s feet are not all properly supported in the same plane as the machine’s feet. When the hold down bolts are tightened, the foot that is “soft” will be pulled down to the base or shim pack. From the time the bolt head makes contact on the foot until the foot reaches the base, the torque required to turn the bolt is less than normal and therefore feels soft. Although the foot is now solidly contacting the supporting structure, the real problem is not been cured. The problem is that for all the feet to be pulled down to the supporting structure there must be a distortion in the machine frame. A more accurate term would be “machine frame distortion”, however this is universally known as “soft foot” and is unlikely to change. The same problem occurs when there is pipe strain, conduit strain, or excessive misalignment. Any of these strains may cause a machine frame distortion.
Machine Frame Distortion
When machine frame distortion occurs, it can cause several problems. The most familiar is how it affects the alignment process. It is virtually impossible to accurately align machine that has uncorrected soft foot. As a machine’s hold down bolts are loosened or tightened, the machine frame distortion changes the shaft position of that unit, which changes its alignment to the other shaft. While misalignment measurements should be taken with bolts tightened, the change from machine frame distortion is not repeatable. This makes the corrections ineffective and precision alignment elusive. Since piping strain and conduit strain also create a machine frame distortion these can be combined with the distortion from soft foot, which complicates the job considerably.
When there is a machine frame distortion it is not easy to predict where that distortion will occur. As the bearing housing is part of the frame, it can be distorted from soft foot. This can pinch the bearing, creating two load zones 180° apart. The distortion also can cock bearing so that it is misaligned to the shaft. This puts considerable strain on any bearing that is not self-aligning. If the bearing at one end of the is machine pulled down it can become misaligned to the bearing at the other end. In an extreme case this could flex the shaft slightly while turning. In addition to putting considerable load on the bearing this situation could eventually cause the shaft crack due to fatigue.
Soft foot also can create a problem in electric motors. Machine frame distortion can distort the stator, which can distort the critical air gap between stator and rotor. This creates electrical vibration at two times line frequency. This problem has become more common in recent years has seen more high-efficiency motors, narrower air gaps and more flexible motor frames.
Foot Related Resonance
Another problem, foot related resonance, is also sometimes referred to as soft foot. To understand a foot related resonance must be clear that the connection of the foot to its supporting structure is not cause of the problem. If there is a resonance somewhere on the machine, it’s base, or piping or anything else connected to the machine, it can be affected by loosening or tightening a machine foot’s hold down bolts. When a hold down bolt is loose and the foot is no longer in contact with the base, stiffness of many parts of the structure can be affected. Even if there is no significant gap under the foot (i.e. 2 mils), the machine is no longer supported on 4 feet, but rather by 3 feet with one part of the machine being cantilevered. The foot itself is most likely not the part that is resonant.
Machines base is also affected. Part of the stiffness of that base came from 4 feet of a machine previously being attached and now only 3 feet. Piping and surrounding parts of the structure can also have their stiffness affected by a foot being tight or loose. For every part that has stiffness changed, the natural frequency is also changed.
The method for finding this situation is my is associated with the term soft foot. Test goes as follows:
- Clear the following procedure with safety and supervision personnel. (At some facilities safety rules may prohibit this procedure, however, no part of this procedure creates any hazardous situation. Explaining the procedure and precautions may allow for a variance in the rules.)
- With the machine full operating condition, measure the vibration at each bearing and in each direction.
- Place the transducer at the point of highest vibration. Record the amplitude and, if possible, the phase at the frequency that appears dominant or unusually high.
- Pick any foot and loosen the hold down bolt just enough so the bolt head does not touch the top of the foot. If the vibration increases to a dangerous level immediately retorque the bolt.
- Observe and record the vibration amplitude and phase.
- Re-torque the bolt, again recording vibration amplitude and phase.
- Repeat procedure for each foot.
If there is a resident part on the machine with all bolts tight, when a bolt is loosened and the natural frequency is changed, it will be less resonant. The amplitude will drop considerably, and the phase will change. This is simply an indication that a resonance is present somewhere on the machine. A more experienced analyst may try loosening various combinations of two bolts at a time, which may have more effect than loosening any one bolt by itself. This could provide clues as to which part of the machine is resonant with all bolts tight. This procedure will not find every potential resonance on a machine, but it can detect most of them. It is therefore highly recommended this procedure be carried out routinely on new installations and on problem machines.
A very useful parameter is watching the phase at the frequency that is suspected of being resonant. If the amplitude changes when a foot is loose or tight, but the phase does not change, then this change is not from a change in resonance. Changing a resonant condition will always change the phase and is a strong indicator of resonance.
Attempting to put shims in the gap under the foot will not cure resonance. This can make it very confusing for those who are trying to eliminate machine frame distortion as any shimming will be ineffective in curing the problem. When that bolt is tight it will once again be on 4 feet, returning to the original stiffness and natural frequency. When seeing how much better the machine runs with one bolt loose and how much worse when it is tight, and no amount of shimming can improve the situation can be tempting to simply leave one bolt loose. This “band-aid” approach will not cure the problem and is usually against procedures. The real solution to the problem is to hunt down the resonant part and add stiffness or mass to change the natural frequency.
Soft foot is extremely common even on new machines and well-made bases. Resonance is more common than many analysts realize. This leaves the possibility of both occurring at the same time. When performing a foot related resonance check, sometimes referred to as a running soft foot check, loosening a bolt can result in machine frame distortion and a changing a resonance at the same time. Very large changes amplitude small gaps under the foot would indicate a resonant situation. In other cases, however, things may not be that simple. Loosening one hold down bolt may cause a machine frame distortion and reduce or increase the resonant condition. Before drawing a conclusion, filling a gap with shim can be attempted. It is important to not over shim under a foot with enough force to cause a soft foot on other feet. If filling the gap reduces the vibration to acceptable levels job is done. If shimming was ineffective or if it helped but the vibration is still too high, the next step would be to hunt down the resonance.
Another potential area of overlap of foot related resonance with machine frame distortion is in the effect of changing the natural frequency by reducing the strain on a part. While a machine supported by 3 feet has a different stiffness than one on 4 feet, a frame under tension has a different stiffness that one without. If there is a machine frame distortion that is relieved when the bolt is loosened, this can also change the stiffness at the same time. Torque can affect the stiffness, as seen by a phase shift while the bolt is turning. To avoid confusion, only compare the phases after they stabilized.
While shimming the gap may eliminate the distortion it may make the alignment worse in the process, assuming a fairly precise alignment was achieved previously (with much difficulty). This could increase vibration at 1x RPM or at the lower harmonics. Another potential confusion can occur on electric motor with a nominal RPM of 3600. Changing the machine frame distortion can change the alignment condition as well as affect the air gap and resulting electrical vibration. Without adequate resolution a change at 2x RPM or at 2xLF may be indistinguishable. At the same time the changing of a resonant condition can amplify or diminish any frequency.
While all these possibilities just nightmare scenarios these can be properly analyzed progressively keeping in mind the various conditions that might be present. Remember that machine frame distortion and foot related resonance are different phenomena, required different methods of detection and different corrections.