CRANE: How to calculate the strengths on hydraulic cylinders? (static + dynamic)

[KIPP]

Hi there. I've got issues. Big ones.
I need to simplify this crane structure, and calculate the strength on cilinder |c1c2| when applying a vertical load on the crane's hook. The purpose of this is to find how much of the strength is absorbed by the first cylinder (and possible the second one |c3c4| as well). I need to devise a heave compensator. If the crane tip has a certain downward acceleration due to ship movement, I need to know how much |c1c2| has to extend in order to compensate the movement and keep the crane tip the same vertical position.

- Points A, B, c1 are fixed.
- The first hydraulic cylinder can rotate around c1
- joint c3 of the first hydraulic cylinder can rotate around B
- c4 rotates around c3
- Distances |DE| and |EF| are variable and controlled by two hydraulic cylinders mounted on top of the crane's boom.

How do I get at this? I've learned how to calculate trusses and beams, but this structure is highly dynamic, complex, and a little beyond my reach at the moment. Is there an easy way to do this? Or a difficult way? By using certain software programs for example... ?

All help is appreciated, even if it is just sending me to the library or whatever. I've been looking through crane books, and books about static & dynamic strength calculations, but I've gotten none the wiser on how to tackle this one...

 

[Pkruse]

This would be a very simple Statics problem if this were a shore based crane. But operating onboard a ship turns it into a dynamics problem. Solutions are pretty much the same, but now you have to account for your accelerations. Do you know what they are?

Do you know how to draw a free body diagram? If not, then you probably would not understand the solution.

Also, loads on your cylinders will change continuously as you operate the crane. Are you only interested in the highest loads?

It may be better for you to go operate the crane and see what the loads are. Put a pressure gage on the dead end of each cylinder, and monitor during the entire operating cycle. If you know the areas of the hydraulic cylinder pistons, then it is easy to convert a pressure into a load.

 

[KIPP]

Hi Pkruse,

I know about free body diagrams and linkages, but I've never had to calculate one this complex, so I'm having some trouble working it out.

Best solution indeed we be to do testing on the crane, or have the company that made the crane send me their test data. The crane manufacturer however is keeping everything confidential and won't give me any data, so I would just try to do calculations with random variables (random boom/cylinder mass/materials), make a spreadsheet, and let someone else enter the actual values.

Just need someone to put me on the right track, especially for the dynamic part :)

Wouter

 

[256bits]

The manufacturer should be able to supply range of movement for each cylinder and maximum load at full extensions so to speak, in other words the things operating charasteristics.

Are you trying to compensate for a ship bobbing up and down in the water and the roll of the ship. I would try to find out how much that is for your ship as a starting point - something sush as 6 feet up and down bob every 30 seconds ( just a guess on my part )
then relate that to how much of extension/ contradtion you cylinder has to do to compensate. From that you can get your hyrdaulic fluid flow rate ( variable of course as the ship's bob is not at a staedy velocity ).

As least from that you have something to start with and you refine the calculations as you go along.

In fact you can do your own calculations of tip/hook mevement as each individual cylinder is utlized if the manufacturrer is not forthcoming.

 

[OldEngr63]

Are points A, B, and c1 truly fixed in space, or are they only fixed relative to the ship? I suspect it is the latter, but this point needs clarification.

Do you have a prescribed sea state and location on the ship for this crane? This will help you estimate the amount of motion, input frequency, etc, if you want to use US Navy documents.

To properly describe the dynamics, you will need dynamic properties (wt, MMOI) for each of the components of the crane. Then you might want to consult the textbook Mechanic of Machines by Doughty for some ideas on modeling and simulation of such a system.

 

[Pkruse]

Cranes like this typically have two failure modes. One is max load & max radius. Another is near minimum radius. I'd do a simple static analysis at both. I'd assume worst case acceleration of .5 g and multiply results by that. This would bound the problem with reasonable conservative estimates until I could get better data. Get a marine load chart from the OEM. Without that it will be illegal for you to use this crane on a ship.

 

[KIPP]

Thank you for your replies everyone. I had a meeting with a company that makes dynamic heave compensation platforms for ships today, and inverse kinematics is indeed the way to go. This translates into a forces, if the weight of the booms and craneload are known (which they aren't, but I can just plug in random variables to do the calculations with). This force can be transformed into a pressure using the area of the piston. Flow can be calculated, so I know whether the ship's hydraulics are sufficient, or an additional hydraulic power pack has to be installed.

I'm going to need to do some work in Maplesim/Matlab however, to calculate the dynamic equation movement matrix, aka the Jacobian (boom angles, crane rotation, and crane jib extension).

A lot of this is pretty new to me, so if anyone has some information regarding inverse kinematics calculations, or modelling in Maplesim...

I will update the topic as I go along, to make sure this problem gets solved. Someone might even benefit from it... if it all checks out that is.

@256bits: I'm going to meet with Bosch Rexroth & another company that makes hydraulic cylinders, give them the outer diameter of the cylinder and ask if they could estimate a maximum load. Or I could just extrapolate the data from catalogues of smaller cylinders.
Ideally, all 6 degrees of motion could be compensate, as I learned today this is possible, however more complex.
Ship motion depends on the ship's size and shape, and can be calculated for simulation using a ship's RAO or Response Amplitude Operator, I'm going to see whether I can get this from the shipyard, or use a similar vessel with an existing RAO. The alternative is measuring it using an MRU, which for example IXSEA and Kongsberg make, but there's no budget on this assignment to do testing, so I'll have to stick to simulations.

@OldEngr63: A, B, and c1 are fixed to the crane's pedestal, which is bolted and welded onto the ship, so they move with the ship.
Sea state has not yet been fixed, but I'll get more on this tomorrow. It's supposed to be a retrofit, so the crane is already in use and on the ship. I've calculated the vector of the crane's tip so that I can track it at all times using sensors that would still have to be built into the cylinders.

@Pkruse: I have a load chart, and reduction factors for different sea states for the crane, but I didn't think this was relevant to the kinematics :)

 

[Pkruse]

It was not relevant. I just mentioned it to make sure you had one. I’ve put a number of cranes onto ships and barges. But I’ve never fooled with a heave compensator.

Loads on the cylinders really are a simple ten minute statics calculation, in the manner I described. My bosses have always insisted on doing the simple hand calculations before going to more sophisticated computer analysis.

Something you may or may not be familiar with is a LMI, Load Moment Indicator. It is a computer that monitors pressure in the cylinders, angles, and boom extension lengths and gives the operator real time feedback as to where he is at any time on the load chart. They have safety interlocks to prevent going outside of the load chart. You might find a convenient way of working that into your design.

If your hydraulic supply is sufficient for normal operations of the crane, then it will also be sufficient for the heave compensator. The HC does nothing more than operate the crane to compensate for heave. What difference does it make if the crane motion is commanded by the operator or the HC?

If you find out that your HC needs more pressure than the system operating pressure of the crane, then be aware that relief valves may very well prevent operation of the HC. If that be the case, the HC will still work, but more slowly.

Also, neither the pump nor the HC will have any effect on the static pressure of the system. That is generated by the load.

I said all that to say that this is not a pressure concern. It is a volume flow rate concern. The crane is only designed to operate so fast, and if you want it to go faster then that is probably not going to happen.

Another thing to be careful of in doing mods like this is to not defeat the purpose of any thermal relief valves on the cylinders. They are often designed into holding valves or counterbalance valves, so it may not be obvious by looking at the schematic that you have a thermal relief. But if you defeat it, then you will burst the cylinder barrels explosively, and spray oil all over the ship. Been there, done that.

Another thing to consider is that it is possible to get serious shock loads into the hydraulic system when you mount a crane onto the ship. This would happen during violent sea conditions, probably when you are not operating the crane. If you combine that with air trapped in the cylinders, then you can get a violent diesel explosion. That can also rupture the barrels with an explosive like sound. Been there, done that. A seven foot split down the side of a boom lift cylinder. Less serious diesel explosions in hydraulic systems will only blow out the seals, or cause heat damage, resulting in a leak. It depends on how much oxygen the trapped air provided to the diesel explosion.

Normally, you don’t have air in the system; but it is easy for a careless mechanic to introduce air during maintenance or repair. This is why all our boom lift systems in such service had an accumulator to function as a shock absorber.

I don’t know how much of this you know or don’t know, or how much of it even applies to your system. I’m just sharing with you some hard learned experience.

 

[FeX32]

inverse kinematics is indeed the way to go

Inverse kinematics would give you the joint rates of the actuators to the joint rate of the end of the crane. This does provide fundamental and important information (like in robotics), however, it seems you need some sort of force analysis. There are several ways to do this. Some are mentioned already.
One other way would be to use the inverse kinematics and further develop it with 2 steps. This can yield a closed form solution to the acceleration analysis. Ideally, this can correlate any dynamic force possible with any given acceleration.

What you need is someone who knows this stuff. It is not something you can learn and play with in a short period of time. Of course, this is if you want a proper analysis or some estimate (static)
Anyways, I could talk about this forever...

 

[KIPP]

I'm afraid I lack the experience, both practical as in design of hydraulics, to be able to work this out properly, covering all the bases. Needless to say I'm in desperate need of people that know their hydraulics and can teach me a thing or two, things that can't be taught through textbooks, or people that know the right textbooks, so I'll probably make some steps toward the industry next week.

The whole heave compensation and crane kinematics thing aside, could one other possible solution to protect the crane be using bigger relief valves? So no position control, just to try to prevent material failure. What exactly happens to the crane if it's subject to sudden shock loads? The relief valve starts operating, but what if it can't swallow the increase of flow? Pressure increases causing the deck to be nicely drenched in a layer of oily sweetness I'm guessing...*

If you have more time, and are interested, I can always email you the crane's hydraulic schematics. I just don't think I'm allowed to throw them on the web for grabs :)

The purpose of the thesis is to come up with several theoretical solutions on how to protect a shipboard crane against dynamic overloads, but I might've gotten sidetracked by the fascinating process of Active Heave Compensation, that I forgot to consider a simpler and more cost-efficient solution

These are the theoretical (not all are equally practical, let alone cheap) solutions i have so far:
- tampering with the current hydraulics of the crane, servovalves, relief valves, ...
- use of an hydraulic accumulator to absorb peaks in oil flow, making it available for the valleys.
- Force control using seastates, measuring the pressures on the cylinders and varying the pressure setpoint on their relief valves according to the current sea states, using load reduction factors I got from Lloyd's register.
- AHC with an MRU, PLC, and closed control loop, using the* crane's kinematics and sensors in the actuators
- buy a bigger crane
- use a lifting system with a winch + A-Frame (f.ex. bosch rexroth's MAHCS), these can be controlled easier

 

[KIPP]

Inverse kinematics would give you the joint rates of the actuators to the joint rate of the end of the crane. This does provide fundamental and important information (like in robotics), however, it seems you need some sort of force analysis. There are several ways to do this. Some are mentioned already.
One other way would be to use the inverse kinematics and further develop it with 2 steps. This can yield a closed form solution to the acceleration analysis. Ideally, this can correlate any dynamic force possible with any given acceleration.

What you need is someone who knows this stuff. It is not something you can learn and play with in a short period of time. Of course, this is if you want a proper analysis or some estimate (static)
Anyways, I could talk about this forever...

I am always more than willing to listen, if you have more information about the acceleration analysis. I know the time frame of this thesis, and the fact that I have very little background means my research can't be very in depth, but I figure the more I can learn from this the better.