Hydraulic circuit design software 9

[frans]

Does anybody in the forum have experience with hydraulic circuit design software. I would like to simulate operating conditions in complex circuits and get results such as flow losses, pressures and temperatures.

 

[frans]

Unfortunately kcj, I don't think it work like that. I am an engineer with a masters degree and have been designing machines for as long as I can remember, so you can say I have a good grounding in just about all aspects of engineering. I understand hydraulic circuits. Unfortunately the devil is in the detail. Those "pitfalls" are things you only understand when you are intimitely involved with something. When you make a real study of it. If that is available in a book, please tell me tyhe title and my problem is solved. If not, you need the time to devote to the particular field of expertize not only theoretically, but you have to catch up with the practical day to day maintenance and problem solving to learn the nuances.

 

[hydromech]

In my experience, the secret with hydraulic systems enginnering is to leave yourself room to change things. All too often designers...that's designers NOT hydraulic systems engineers, expect their designs to leap off the drawing board and start working...fat chance, it just doesn't work that way.

The mantra chanted by many many production managers all over the world is "more power, more speed". Too many hydraulically operated machines are installed, commissioned and then abused by the customers.

Counteless times I have commissioned a system, trimmed it and got it working perfectly. 1 month later, the customers on the phone..."this machine dont work". I go back to machine to find the usual problems...compensators set above the system relief or the relief valve so high that stuffed the pump or split a manifold.

"Why have you done this?" I ask..."we wanted it to go faster" they say.

I always try to leave room for change...it is inevitably required.

My opinion on hydraulic design sofware is that it's O.K to a point. You feed information into assuming that all the information is correct...speeds...inertia...power available...torque etc. What happens if one of the paramaters is wrong? The software can't tell you. Everyone else will though...when the machine doesn't work.

All the software in the world cannot replace experiance. The best experiance is when things go wrong. Make a note of what happened and why and don't do it again.

Hydromech
Hydraulic Systems Engineer

 

[kcj]

frans:
I think we are saying the same thing. My point is that some managers (here but also at many other companies I know of) treat fluid power, and design in general, as a black box commodity that can be sourced from WalMart, kMart, or Target, whoever has the best price. As long as the project is defined in costs and schedules, the situation is 'managed' without understanding the nuances and details.

Experienced people are not respected because the outside expert has more aura and a better presentation of the latest technology. The major design decisions are frozen, the project comes back in house, and the people with the day to day engineering or operations knowledge are involved only in paperwork processing, until startup or support when it is too late to change key decisions. The old 'xx% of product life cycle costs are determined in the first 10% of decisions' or something.

I have a couple degrees, but mainly 30 years of learning and scars. Sounds similar to your background. My pipe dream, and what I was trying to say, is that all of that intimate knowledge of the details should be brought into a project early on. Use the outside vendors and experts who are up to date on the features of the latest product or technology in general. They are the people who may know better what 'can' be done. Add the experienced inside people who know what is the goal, what are we really trying to accomplish. That hybrid of closeup experience and wider view vision could be very efficient and productive.

kcj

 

[PNachtwey]

kcj, one of the biggest problems we have is getting the OEM to take ownership of the machines he is selling. The OEMs claim they are too busy to watch what we are doing and then they need us back to help with the next machine. On top of that, they don't understand what is happening in the black boxes to explain how the system works to the end customer. We do offer training courses for the motion control part. Most companies do the right thing and send there people to be trained. It is the other %10 that is a problem.

hydromech, I see the same problem all the time. If the system is a sinusoidal test system, how much more power is required to increase the power by a factor of two?
Knowing the answer to this should keep people from trying to go faster than the designed speed. Hydraulics can be fairly linear if the pressures inside the cylinder stay in the middle third of the pressure range. Increasing the speed causes the pressures in the cylinder to swing out side of the this 'almost linear band' and the now the system is much harder to control.

 

[frans]

Hydromech, why should I not expect a hydraulic system to work straight off, maybe with one or two adjustments? Relief pressures can be set beforehand. Flow regulators can be set before installation. If I spec the functional requirements to the satisfaction of the hydraulic designer he should be able to deliver a working system, or else it is either black art or he has not paid enough attention to detail.

I agree with allowing flexibility in designs. I agree the functional requirements has to be specified properly.

I do all this and I still get systems that do not work.

 

[PNachtwey]

"I do all this and I still get systems that do not work."

Why not? If you have an data acquisition system you can see where your assumption were incorrect and learn to make better assumption or calculations. Data acquisition should be a new thread.

I have a system in the warehouse where I have position feedback and pressure sensors on all the ports of the servo valve. I use this to verify my models. I can acquire up to 16 data streams with 8 of those being external data at periods as short as .5 milliseconds.

 

[frans]

Not all hydraulic systems uses data acquisition. It is nice if cost is not a problem, and I do have those, but not always. Secondly, how does that absolve hydraulic designers from coming up with proper designs in the first place?

 

[PNachtwey]

"Not all hydraulic systems uses data acquisition. It is nice if cost is not a problem, and I do have those, but not always."

The data acquistion tools are not part of any one particular project. After you get the project installed then you take the data acquistion sensors and electronics with you. This way you only pay for the sensors, AtoD converters and software once.

"Secondly, how does that absolve hydraulic designers from coming up with proper designs in the first place?"

I use my simulator. As mentioned above, there are packages like the Automation Studio. I saw the Automation Studio at the last IFPE in March. I was impressed. However, I don't think the simulations are as good as they should be, but I can see they are on the right track. If Automation Studio can do what my simulator can do then it will be a must buy. By the way, Automatin Studio can do many things my simulator can't do and it is a polished package whereas mine is not.

When I was at the IFPE show, I gave the specifications to my hydraulic system to the Fanuc ( the compay that sells Autmation Studio ) sales manager. He quickly entered the data. It took about 10-15 minutes. When the simulation was started I could see the pressures, speed and velocity change. However, there was no way to see the values at any time in the simulation so I could not compare his results with my results. It is this lack of quantitive data that keeps me from giving Automation Studio an unreserved must buy.

 

[hydromech]

Frans...

Theoretical spec from a hydraulic systems engineer to a customer.

1. The power available to drive the hydraulic system will be infinite.

2. The noise emission might not be lower than 90 dBA.

3. The fluid used will be a standard 32 CST mineral oil.

4. All actuators will be located within 5 metres of the power unit.

5. All of your inertia calculations will be accurate to within 1 mtr/sec/sec.

6. There will be no limits to the size of the pipework and the pipework installation will not be compromised and will not deviate from our requirements.

7. You will surrender the right to change any part of the design of the machine once we have started to design the hydraulic system.

8. The fluid will be clean to ISO 4406 level 18/15/12.

9. Once the machine is commissioned by our engineers, all valve and pump adjustments will be sealed to prevent tampering by unqualified persons.

10. You will deal with third party warranty claims. If any part of the hydraulic system under performs due to efficiency problems caused by the hydraulic OEM's(pump or valve supplier)inability to publish accurate performance data we will not be liable.

11. We will charge you a fair agreeable amount for commissioning the system and training your staff on hydraulic systems maintenance.

Failure to comply with the above will void the warranty.


Lets face it Frans...this aint gunna happen.

In lots of cases the specification placed upon the hydraulic engineers requires a specific solution. That specific solution is shall we say..."the ideal".

The hydraulic engineers pass "the ideal" system solution to the the customer or to the mechanical designers or to the electrical department.

Hydraulic systems engineers are a pragmatic collection of people, "the ideal" solution does consider every one else...it has to. Unfortunately, that concideration is not often reciprocated. The usual replies are "You can't put that there" or "that pipe will have to go around the machine" or " you cant have a motor that big...we haven't left enough room for you.

Already the hydraulic systems engineers are on the back foot...compromised...playing catch up and these are the known problems, they don't include cock ups and design failures.

Many many overall design failures manifest themselves in the performance of the hydraulic system. All the fingers point at the hydraulic systems engineers.

Frans...it should go right and it usually does. Hydraulic systems engineering is not a black art and my attention to detail is often offset other peoples lack of knowledge and experience.

Hydraulically operated machines cannot be delivered preset. I have seen brand new presses snapped clean in halve because some left a relief valve at a test pressure. I have seen the machining head on a brandnew boring machine fly clean off because the flow control valves was "preset"...fully open! It is best practice and basic common sense to shut all flow control valves and back off all relief valves to the lowest possible pressure bofore shipping the machine.

Hydraulc systems engineers : "Would you like us to commission the machine for you?"

Customer : "Nah it's O.K...I mean how difficlut can it be...it only turning a few knobs...isn't it?

1 month later...

Customer : "this machine wont work...there's something wrong with the hydraulics...we've had a got at it but its just getting worse...WARRANTY!"

Everyone has a different experience...this is just mine.

 

[frans]

You all make fair points. However, in my case the standard of expertise, design skill and professionalism offered by the hydraulics suppliers, and I have tried all the majors, who all offer to do the design for me, is not up to scratch. They want to sell components, they do not want to support or be held accountable. In general I do not find this with electrical control system suppliers. Or rather, I can easily pinpoint those indivicuals with the knowledge and skills to deliver what I need. Not with hydraulics.

Unlike the hydraulic suppliers I am accountable to my customers. Therefore I am moving to do my own designs or at very least being able to analyze what I am being offered. When i do mechanical machine design I calculate stresses and and strains and displacements and moments and forces etc, but I have never seen a hydraulics expert calculate in detail the flow losses, and heat build-up and dynamic response etc. Yes, I am a "it works and stay working after commissioning" type guy. I fully accept there are individuals out there capable of doing so. Only I have never come accross them.

Patchey, I take your point on instrumentation and data acquisition. I gather you feel automation studio is not the final answer? Is the simulated data calculated, dispayed and then discarded - not stored in a file somewhere accessible?

Hydromech, I take your point on commissioning. See, I am learning.

 

[hydromech]

Frans...

You are right about the big OEM's.

I have had experiences with all of them. Parker..Vickers...Rexroth...Denison. They are the same.

At the front end the sales people...who generally don't know beans about what they are selling or how to correctly apply what they sell.

Behind them are the product engineers...kept so far away from the customer...they couldn't give a toss.

I have been working in the fluid power business for 17 years and every day is still a school day. If you ever hear anyone say they know everything about hydraulics...run away from them...they are dangerous.

Lets not forget the name of this web site Eng-tips.com...this thread started as a question about design software and look where we are now...? I though the idea was to help each other out.

There are lots of people out there giving hydraulic a bad name...I want to try and put a few things straight.

If you need help...ask. Thats what we are here for. You will get an honset answers for free. There aren't many places that can offer that.

 

[EdDanzer]

Depending on one manufacture to design and supply a complex system that will work flawlessly is not going to happen in any area, hydraulic, electric, electronic control, machining, fabrication, ect.
No one manufacture supplies best of breed in all hydraulic components, yet they will use marginal components to be sole supplier. I use components from both Sun and HydraForce even though they make products that do the same function. The reason is, each company has their own design for different valves and one may out perform the other is specific applications, the only way to find out is to test. If simple projects are failing, then the information given to the supplier must be suspect, not everyone can screw up every project.
We just went through a problem with a machine tool where the ladder logic and user interface would not work correctly or the same every time you pushed a button.
Simulation software will help, but if every detail is not set and followed, the results will be different. The size and in some cases changing the ID of a tube or the manufacture of the fitting or hose end will compromise the system. Long runs of pipe or hose can have ill effects. Then some problems are freak. I designed a system using a HydraForce CV08-21 check valve. The flow was high enough that the back pressure was a little to high, so I installed a CV10-20. This valve caused a malfunction. After much testing and instrumentation, we found the designs were different and the CV10 could float open during a pressure balance condition. Some things in life, are just learned the hard way, trial and error. No software could simulate these valves to determine if this condition existed as HydraForce didn’t know this could happen. Obtaining values for hydraulic valves would require testing valves constructed with minimum, average, and maximum tolerance, then using that range for simulation. The next problem would be hose, hose ends and fittings.

 

[PNachtwey]

"Obtaining values for hydraulic valves would require testing valves constructed with minimum, average, and maximum tolerance, then using that range for simulation. The next problem would be hose, hose ends and fittings."

We shouldn't have to do the testing though. The manufacturers should supply this data. This is the biggest problem I see with simulators or even doing the calculations manually. There isn't enough accruate data.

I want to know what the capacitance of pipe and hose is. I want to know the equation for the spool area on a Bosch NG16 with a nic spool. I want to know how to model the compensator on the pump. Is the response of the compensator exponential or is it a combination of exponential and slew rate limited? This information is not available.

Frans, Automation Studio looks close, but ready for use as a design tool yet. Fanuc is pushing Automation Studio for class room or training purposes. I will dust off an example of one my simulations. I have a .pdf site where I can post it later.

 

[PNachtwey]

Here is my attempt at a simulator.
ftp://ftp.deltacompsys.com/public/PDF/Mathcad%20-%20Hydsim-ps.pdf

I, actually my customers, do a lot of hydraulic servo control. This is a Mathcad worksheet that I use to sanity check designs. I can see how the pressures will change as a function of time. I can also check to see if the actuator will accelerate and decelerate properly. In this example the pressure graph on page 7 is most important. I can see the pressures are getting close the system pressure and the blind end pressure is getting close to cavitating. Also, notice the ringing at the end of motion. This system is not very stiff. This system did not work well. It had to be re designed to use 3.25 in diameter cylinders to get the required stiffness. This was for a flying shear and it must track within .010 inches. If the cylinder is not stiff the 'slinky' effect can be greater than .010 inches.

This simulator is solving a system of differential equations using iterations of increments of 125 microsecond. I have more information on how the simulator works here.
ftp://ftp.deltacompsys.com/public/PDF/Mathcad%20-%20HydSim.pdf
It use a numerical integration technique call advanced Euler. It is being rewritten to use Mathcad 12 and the Runge Kutta method.

The program is actually a little more sophisticated than the documention says it is. I have added pump flow and accumulator volume to the simulation. What is neat is that I can try different things like running the piston to the end of the cylinder to see how the pressures change. I can hang a mass veritically and see the actuator cavitate on the cap side.

 

[frans]

I agree fully with PNachtwey: "We shouldn't have to do the testing though. The manufacturers should supply this data. This is the biggest problem I see with simulators or even doing the calculations manually. There isn't enough accruate data."

Maybe one of you guys specializing in hydraulics should compile a standardized component information framework covering the test and specification information that should idealy be made available by components suppliers.

You may even be able to set up a site such as Tomshardware to compare different brands on standard tests and make money out of it.

Patchey, your simulation is impressive. Again, you may find a market if you add enough block components so users can build systems. I build complex spreadsheets for machine design because I know the stuff so well, but it would take me far too long to set up hydraulic simulations like you have done. Hence my search for something off the shelf - I do not develop my own FEA software for instance, even though I can do it from first principals! Used to do that years ago to be fair. Just too busy now.

 

[HCMechSys]

Has anyone tried the free PCBathFP software for hydrualic circuit simulation?

I teach basic hydraulics and have gotten the students to use it to creat basic hydrualic circuit diagrams. I'm thinking of incorporating some basic simulations as well into the course. As someone previously posted, I believe simulation can only do some much before you have to build prototypes and try some experiments. But I thought an exposure to simulation is important as well.

Any thoughts on the software greatly appreciated. It can be downloaded from:

http://www.bath.ac.uk/mech-eng/ptmc/pc_bathfp/pc_bathfp.html

 

[frans]

I tried it and it works well, but is limited in the components that are available. To get components added you have to contact the University and I can just imagine how long it will take our theoretical cousins to add just one component - so as a practical tool unfortunately I think of limited use.

 

[HCMechSys]

frans,

Yes, I've tried the "support" and its definitely not the fastest. But for a free tool, that's to be expected.

Have you tried the simulation capability? I've tried a simple example involving an unloaded hydraulic motor, pressure-compensated throttling valve and counterbalance valve and have had problems getting the results to agree with measurements from a corresponding circuit I put together using a Parker training station. The program seems to play with the volumetric and mechanical efficiencies and prevents me from adjusting things to get the results to agree. I've tried the "support" and they've never gotten back to me.

I was curious if this sort of thing is common with other tools like Easy5, Automation Studio, etc or is this a quirk of the BathFP program. Any thoughts or examples to try most welcome....

 

[HCMechSys]

I thought it might be a useful exercise to try to run PNachtwey's servo-valve example using the PCBathFP software.

A summary of the simulation and comparison of results is at:

http://ca.geocities.com/hchesser@rogers.com/Ex-Eng-Tips.pdf

I found that the analysis basically confirmed what PNachtwey was saying the system is not stiff enough to drive the heavy load. However I found that the PCBathFP software did not let me set the initial conditions for the accumulator and line pressures so I was not able to replicate the simulation. I've e-mailed the Bath people to see what can be done. In the meantime, I think I've thought of a way of setting up the right initial conditions.

I did have a question about the initial line pressures - why were these particular equations/pressures chosen?

 

[PNachtwey]

Wow, nice effort. Thanks. I haven't had time to examine the BathFP software.
I think there are a number of things that make the two simulations different.

1. You should have used 35 bar instead of 70 part for rated flow for each land of the spool. Notice my ValvePSID/2.

2. You used linear instead of cosine ramps. Linear ramps have high frequency components that will induce oscillations. Cosine ramps have only the one frequency component.

3. Some how you got the actuator to retract too far and back up against the cap end of the cylinder at 1.35 seconds. The BathFP software does exactly what I would expect it to. My simulation did not hit the end of the cylinder.

"I did have a question about the initial line pressures - why were these particular equations/pressures chosen?"

Are you are referring to Pa0 and Pb0? If the initial velocity and acceleration are 0 then there must not be a net force on the piston. Just to the right of where I set Pa0 and Pb0 is calculate the force on each side of the piston to make sure the force on both sides of the piston are equal. I can also use this simulation to handle vertical loads. In this case the load force on the rod will be non zero.

If you are referring to P0 ( pre-charge pressure ) or P1 ( system pressure ), the system pressure be high to keep the system as linear as possible for the motion controller. The pre-charge pressure is high relative to most hydraulic systems. I like to keep the system pressure as constant as possible to make tuning the motion controllers easy and reduce following errors.

HCMechSys, I think you should be able to fix items 1 and 3 fairly easily. It would be nice to rerun the BathFP simulation. I am also interested in the feedback you get from BathFP. I will work on making my simulation metric friendly.