Leveling a large load with hydraulic supports


Thread Starter

Paul Richmond

The trailer mounted turbine unit must be operated only in a true level state. There are at least two sets of inclinometers that measure the pitch and roll of the trailer - one set at each end, as the unit can 'bend' under it's weight, hence the need for eight supports. The supports can be hydraulically extended and retracted, but actual extension position is not known. Only limit switches to indicate max or min extension. I had thought of a PID loop for each leg, but since each leg's position will affect the pitch and roll to at least one other leg and with only two sets of inclinometers, this can't work - not in it's purest form at least. Any thoughts as how to approach this problem ?
Paul: That huge turbine I would like to see it (maybe in picture). The float should be leveled before departure and each one of the eight pistons should apply equal amount of pressure under the turbine, in order that she is not subjected to bending moment. That requires load cells telling the amount of pressure is applied to each piston. Each pressure applied must now agree with the mathematical "pressure model" the engineering dep't should have (it's the most important parameter. Otherwise, you would be working blind. At this point, your company should know if a spring bed is required to depose the turbine on the float. A single phydraulic pump would feed all eight pistons, each hydraulic line to the pistons being equiped with laboratory type accurate manometers (big dial). Each hydraulic line having a needle valve to let "pressure in" and a needle valve to let "pressure out". Set the pressure points in turn, little at the time, constantly reading the master model. Now, the turbine seats on her own nest, in perfect equilibrium. If the float has rigid enough frame and very many tires (I have 48 tires float), then move. Not too fast. The float is the all key. What about if you lift one end of the float ? Will it create bending moment to the turbine ? NO. Then you do not need inclinometers. That turbine is certainly not designed to stand up, but she can support inclinations (+/-) during the trip. May be I'm wrong in that assumption. ??? Don't forget to use certified hydraulic fluid, having negligible compressibility at the pressure it will operate. Use the very best workmanship in the tubing/fittings. The entire hydraulic system must be perfect, and trustable. It may be wise monitoring the system for few days before "le grand depart"; and keep monitoring each pressure point during the journey. Paul, send me e-mail after arrival. In the event the turbine must stay horizontal while the road varies in inclination: then the system requires pressure compensation, distributed linearly as per the variations in inclinatiion. The strategy is a feedforward "Master loop" feeding the eight pistons. A reliable multiloop intelligence system must be installed on board. [may be National Instruments ?] Then, the hydraulic circuit must have recirculation to let the fluid in and out. This requires the needle valves become "control vaves" [check Swagelock, check deeply with Parker] Thanks Paul for offering me an interesting breakfast chat.

Adolfo Jimmy Saldivias

Paul: What do you mean by level state? I think one of the most important concerns about your application would be that all 8 supports would be supporting the same load. If I am right, you'd better use loadcells under each support to monitor the load. Jimmy Saldivias TECSIM Phone: 591-4-523438 Fax: 591-4-523413 http://tecsim.trading.net

Richard King

I can see the concern, but I am sure all eight supports won't have the same, also I do not think the loads will be constant at each support. Load Cells might be useful to monitor/control the stress on the trailer frame between each support. Richard King Alesa Alusuisse Engineering Australia Pty Limited. Phone: +61 7 3218 3555 Fax : +61 7 3236 0155 Email: [email protected] Web : http://www.alesa.com.au
First of all, please don't yell at me for this explanation. I'm not able to take the time to solve the whole problem, I just want to give you some things to think about ;^) Since you don't have feedback from each cylinder, I would look at using hydraulic flow dividers to make sure that each cylinder gets the same flow. As long as you bring the cylinders home, this should work (unless the cylinders are really leaky). If you just maintain constant pressure to each cylinder, the flow will go to the one with the least resistance. The movement of the cylinders is determined by volume, not pressure. Flow splitters are relatively common hydraulic devices, although you would need a lot of them to make it work right. Here is an example diagram of how you would use flow splitters with eight cylinders. Each "+" is a splitter. This would raise and lower the platform (all cylinders together), assuming it were already level: ---- cyl 1 | ----+ | | | ---- cyl 2 ----+ | | ---- cyl 3 | | | | ----+ | | | ---- cyl 4 pump ---+ | ---- cyl 5 | | | ----+ | | | | | ---- cyl 6 ----+ | ---- cyl 7 | | ----+ | ---- cyl 8 What you may be able to do with sufficient valving and splitters is to divide the platform up into four different sections, with two cylinders for each section. Then you could do a "bang-bang" control loop with the inclinometers for the two different axes. Regards, Willy Smith Numatics, Inc. Costa Rica

Paul Richmond

In 'theory' I would agree that you would expect all supports to carry the same load. Our actual need however, is to ensure the whole trailer is horizontal with respect to gravity prior to any turbine operation - the orientation can't be changed during turbine running. Reasons for doing this include balanced loading on bearings etc. and yes, potentially the load will change during turbine operation, but since we are unable to correct for this during this period, we can only monitor the shift in position and alarm appropriately. I've had a lot of good input as a result of my original request and it's given me good 'ammunition' to return to the trailer manaufacturer/turbine packager and point out the failings in the original design. They have as of today agreed to only use four (4) - rather large supports as opposed to the original eight. The inclinometers will remain and leg extension position will be monitored for each support. Hence, the required solution is now very much simplified. Thanks to all for the help and suggestions, it's appreciated. Paul