Reuben Hale, P.E.
Ph: (510) 507-1300
Much of the vibration problems in heavy industry involve rotating equipment of some kind, from power generation, mining, paper mills, agriculture, to food processing. The dynamic interaction of the rotating equipment and the surrounding structures present challenging problems that are hard to debug without the help of expert vibration and structural dynamics analysis. Sorting out cause and effect in the structural dynamics puzzle can help the engineering team target the most efficient approach to solving the vibration issue.
Below are some examples to illustrate the variety of projects we work on.
Paper Mill Structural Testing and Simulation - Paper mills involve large dynamics forces at multiple speeds. We have worked on paper mill projects that involve the dynamics of elevated structures.
|Paper Mill Structural Dynamics|
A quality control lab was experiencing problems with a very sensitive scale for weighing paper samples. We came up with options for modifying the structure, the scale frame, but perhaps the most cost effective option was a new measurement algorithm we developed for use in sampling the weight to better reject noise due to paper winder vibration. We also recently worked on a project to design a new elevated platform for the motors and gearboxes used in driving the wet end of a large paper mill. To do this we needed to perform extensive dynamic testing and vibration characterization of the existing structure and drive system, and then used those force estimates, geometries, and excitation frequencies as inputs to Finite Element Model. We used this FEA model to simulate various construction options for the contractor.
Crankshaft Failure & Engine/Generator Dynamics- We performed Modal Testing, as well as Operating Deflection Shape Testing (ODS), on the 6 engine generator sets used to power a semi-submersible-dynamic-positioning-vessel. This ship is used as a work platform to service oil rigs in the Gulf of Mexico. Two of the six engines suffered from repeated crank shaft failures, even after expert inspections and rebuilds, at an enormous cost.
|Crankshaft and Counter Weights||Sensors on Engine Block|
We flew out to the ship and performed vibration characterization and modal testing to see which engines had the highest levels of vibration, and also performed modal and resonance testing of the isolation systems to understand why these two engines were having failures while the others were not. Differences in the engine isolation system dynamics explained the higher levels of vibration for the 4 engines that were not having failures, but did not correlate with, or explain, the crankshaft failures that were peculiar to 2 of the engines. This was a surprising yet very helpful result that suggest that it was not the absolute vibration of the engine that was the cause of the failure. Could there be a difference in the deformation of the engine blocks that would explain why the two engines had repeated failures while the others did not?
It answer this question prompted by our previous results, we looked carefully at our operating defection shape analysis (ODS) and compared the deformation shapes of the engines that did not have crank shaft failures to those that did have crank shaft failures. Careful testing and analysis showed that the operating deformation shape of the two engines that failed was characterized by deformation of the engine block in two axes, while the engines with much greater vibration, but with no failures, showed typical engine block deformation in only one axis.
We used concepts of basic physics to show that a properly balance crankshaft should counter the inertial bending moments of the pistons in one axis but not the other. An improperly balanced crankshaft would not balance in either axis and thus we suggested that the crankshafts in the engines with the failures were not the same as the crankshafts in the engines without the failures.
|Semi Submersible Dynamics Positioning Vessel|
This came as a surprise since the crankshafts had been inspected and replaced by experts multiple times. However, this was indeed the case. They had repeatedly put in the wrong counter weights on the crankshafts after the first failure which perpetuated the problem. Solution in hand, we swaggered off into the sunset.
|Engine/Gen Set of Oil Drilling Vessel|
Fan Blade Testing and Simulation - We have instrumented cast fan blades on several occasions with strain gages to look at fatigue and failure issues. The strain measurements help to validate the dynamics and static stress predictions of a computer FEA model of the fan blades. The model subsequently refined, and is then used to estimate the benefits of various design changes.
|Fan Blade FEA|
We use state of the art strain gage instrumentation that is built to withstand many Gs of centripetal acceleration while mounted on the rotating structure. We often use either slip rings, or in some instances, telemetry, to carry the strain signals out to our equipment. What makes our services so unique is that on every project we are involved with we are aware of the larger picture of the many factors affecting the structural dynamics of the equipment we are working on. Almost always, our level of experience is key to saving our client crucial time and effort.
|Fan Blade Test Data|
For example, while instrumenting and measuring the strains on the cast 8ft fan blades we noticed a subtle structural detail that indicated that the test fixture was poorly mounted to the concrete slab floor. They had used the same mounting procedure for years, but in this case, with this particular new design, with an unusually small clearance between the fan shroud and the tip of the blade, the loss of stiffness between the shroud and floor was critical. It was thought that instead of the blade exploding due to a casting defect, the blade may have contacted the fan shroud and caused the catastrophic failure.
The way the shroud was mounted would cause random variations in the stiffness of the connection of the shroud to the floor every time it was reassembled and confounded future testing results. This would have created frustration for both our client and their customer (not to mention the life threatening danger of the shattered cast aluminum pieces that left fist sized holes in the tin roof of the warehouse – once was enough). If we had focused solely on a set of strain measurements we would have missed this larger, more critical, issue. Thus, while we were hired to provide strain measurements, we also provided a very crucial review of their test setup as part of our work.
Power Generation Dynamic Testing and Characterization - When a very complex structure is thought to be the cause of vibration to an adjacent structure, there are often many interconnecting elements/systems that are suspect. To further complicate the problem, there are also often multiple suspected sources of the disturbance.
We will often start by forming a number of likely hypothesis based on our initial analysis of the system and discussions with the client. We will then suggest a test plan to prove of disprove these hypothesise by making measurements of the system response to ambient, and known stimulus.
Stain, or stretch, in a structural element is one way to show that force is being transmitted along a particular piece of the structure. We can then ask the questions:
- If there is force transmitted through a structural member does its magnitude explain the resulting response on the receiving structure?
- Does the frequency content of the force match the disturbance frequency on the receiving structure?
These are some of the questions we can answer using strain gage testing to prove of disprove a hypothesis. This often involves crawling around on beams in dark corners of a noisy plant to install strain gages in hard to reach places. Or climbing up 200ft of ladders to test an oil filter mount in a Wind Turbine. Often, the ability to positively identify a force transmission path makes this effort worthwhile.
For example, we were called in to help identify the cause of vibration to the control room of a newly constructed 120 Mega-Watt power plant. The vibration in the control room was so bad that the control personnel could not read the monitors. The was also a stairway that was unusable due to high vibration levels, and a few stalls in the mens bathroom that were at uncomfortable levels as well. We designed a quick bracing fix for the monitors to make the situation bearable and went to work to characterize the problem. We started with an Operating Deflection Shape Analysis that showed that the turbine was axially coupled to the control room building and not shaking at the spinning speed of the turbine but at a lower frequency that appeared to be related to the power load.
We then looked at the structural connections of the many large pipes that connect the adjacent building to the turbine. It was thought that one, or more, of these large pipes may be transmitting the forces to the adjacent building caused by a flow induced instability in the supply steam. Additionally, based on our acceleration measurements on the receiving control room floor, we calculated that it would take about 50,000 lbs to excite the control room structure to the levels of measured vibration. Thus, we had both a disturbance frequency to look for, as well as a total force level to search for in these various connections.
|Strain and Acceleration Measurements on Pipe Supports|
in Search of Large Transmitted Forces Between Pedestal and Building
We installed strain gages to estimate the forces passing though each pipe support that was suspected to be part of the load path. We were able to show that there were no large forces transmitted though the pipes. This left the final hypothesis of the seize up expansion plates around the turbine. These expansion plates were thought to be seized up due a difference in thermal expansion between the slow thermal time constant of the massive concrete turbine pedestal and the relatively low thermal mass of the steel building structure. The building had been constructed during the winter and a calculation of the temperature of the concrete turbine pedestal suggested it was still far thermal equilibrium. A calculation of the total maximum frictional force of the expansion plates showed that the expansion plates were a possible force transmission path. Rather than instrumenting these plates it was easy enough to jack up the expansion plates to remove the force path. The vibration problem was solved. Thus, we used the stain gage testing narrow down the list of suspects and find the offending structural load path.
We succeeded where many power generation specialists had failed due to our decades of hard earned experience and structural dynamics expertise.
- Heavy Industry Consulting Topics Include:
Overview, Manufacturing & Plant Process, Rotating Equipment, Vibration Induced Fatigue, Power Generation