Sign in to follow this  
Followers 0
mracer

Porject gone bad

18 posts in this topic

I have a project that has become a real problem and I would like to ask for your input. The project is a hydraulic pump with a proportional pressure control and proportional position control valve’s. These valves will be controlling a 120” cylinder that need to be able to move at a max of 120”/sec. A ML1500 was used in the spec for control with a string pot being used for position of the cylinder. I have more spec’s for the hole project if someone is interested in working on this. We had two different engineering groups bid on this. The bid was for assembly of the control panel and programming for the control to meet the spec’s given. The bids where both vary close and we went with the one that seemed to have the best handle on what we where asking for. Not the low bid. 17'000 original quote. We are now at a point where the engineering firm has sent us a bill for $13’000 in “extra work” that would seem to be over run, over what was bid. We did not approve this over run (extra work) in any way shape or form, as was clearly stated in the P.O. contract. They have not obtained the level of control that was on the bid and are telling us that an additional $10’000 will also be needed for changes to the control package. I now have the task of ether finding someone to complete this job or paying the first firm the extra and hope that they can complete the job with some sort of satisfaction. How should I approach this? Thanks,

Share this post


Link to post
Share on other sites
You are raising one of the oldest and most lamented problems of the automation indiustry and I have been on both sides of the arguement. Customers always say that vendors add charges to get what I bought in the first place and vendors tell stories of customers that change their mind every time the sun rises or the wind changes directions. My first thoguht is what is your history with the current firm. Do they have a long history of good service and just missed the estimate on this one or are they a new vendor trying to earn your trust and business. If they ahve a good track record and you want to do more business with them I'd propose metting them partway on the overages and seeing if they can trim some fat from their $10K finish estimate. After all nobody wants to go to court over failure to meet contract. Been there and only the lawyers make money from that path. You also don't mention if this $13K overrun number is their first mention of this number or if it is the result of a long series of negotiations. It could be as simple as they fat fingered their estimate and this number is a talking point and theya re open to a counter proposal. No matter what you do, you will ahve to provide much closer management of this project to finish it with any thing resembling a decent ROI.

Share this post


Link to post
Share on other sites
It all depends on the 'needed changes'. However, since the system integrator seemed to be clueless in the beginning I wouldn't hold much hope for a better solution. I would find someone that can design the hydraulic and mechanical part of hydraulic servo systems and make sure the physical system can do what you want. Then the control will be much easier but it will not be easy. I would bet the true fault lies with the hydraulic design. Very few can design a hydraulic servo system correctly. There are some key questions I would have. The the answer, or lack thereof, to my first question is key. What is the natural frequency of the system? If you don't know this then stop and find someone who does. What is the diameter of the cylinder and the rod size? What is the mass or load? What is the size of the valve. What valve is it exactly. Is it a servo qualify valve? Where is the valve mounted? The cylinder must travel at 120 inches per second. That is very difficult to do. Is there any position specifications? Time specifications? Pressure specifications? Why would there be pressure specifications on something that goes this fast. This isn't a die casting machine is it? It would be a big one. How many cycles per minute? I have seen PLC/drive people work on motion systems for months not realizing that they could never make the sysetm run right because they didn't understand what was wrong. A common hydraulic design mistake is to size the cylinder diameter too small. This makes is easier to meet the flow requirements but it may be difficult to get the load up to speed. Stopping while extending may be harder yet because there is less surface area to slow the load down. Finally, be prepared to spend a lot more money if the hydraulic design is screwed up. What you are trying to do isn't easy the first time.

Share this post


Link to post
Share on other sites
We have all been stung by project cost problems, similar to this, in our careers. Here are some issues/questions to address... 1. Is the supplier/vendor billing you or giving you a quotation for what the over charges would be? If they are billing you, then they proceeded to correct the issue and the work is complete. (determines if supplier fixed the problem and are forcing you to pay for the work or if they want permission to resolve) 2. Has this supplier/vendor done project work for you before? If so, what has been their history with you (on-time, on-budget, etc.)? (builds the case for or against supplier proven capability) 3. Do you have any similar systems as this in your facility that you can use as reference? If so, how does this supplier/vendor solution fare to what has already been proven in your facility? (sets the stage for known solutions) 4. Ask this supplier/vendor have they provided any similar solutions for other customers before (sets the stage for capability or not) Your options are to pay their charges and go overbudget (and get spanked by the project cops), negotiate with the supplier to trim their quotation, have the supplier finish what they contracturally obliged to complete and death with the aftermath later, or pull the plug and pay someone else to make it work properly (or do it yourself). Many of us have been all of these situations before (for some, many occasions so). Many of us have heard stories and have run into suppliers/vendors who pad their sales dockets with over charges. One such vendor I know of is well know for coming in low-bidder, but they make their profit back after the first over-charge. My goal, as the project engineer, is to write a contract to cover my employer's tail and keep from having over-charges that blow my budget. If the supplier goofed up and has to fix their mistake, your choices are to pay the full charges, pay nothing or split the charges. When it comes to an honest mistake on anyone's part, splitting charges is acceptable (nobody wants someone to go out of business, even if they or you screwed up). Things to learn from this situation... 1. Did your scope of work have holes in it that allowed for this to happen? 2. Did this supplier/vendor understand the scope of work? 3. Does this supplier/vendor have the qualifications and project history to pull this project off? I have always taken the outcomes of past projects and used those events to drive the development of a base project specification to be used for future projects. Your tail is on the line, not the supplier/vendor. It is business, not a charity.

Share this post


Link to post
Share on other sites
As a guess, given the bid price, the vendor was not supplying any hydraulic equipment. Who did the hydraulic design and supply? Who decided that a ML1500 and a "string pot" were appropriate? Was the panel builder given a spec asking for a panel containing specified components that others had determined to be adequate for the specified "level of control" of the hydraulic system designed and built by others? You should seriously address all of Peter's questions - maybe even contract him to vet your system (if he's willing).

Share this post


Link to post
Share on other sites
The specs for the hydraulic system where given to each of them before they did any bidding. I was told the ML1500 would have the ability to do the job BY them. I gave them a set of abilities that the system needed to meet. I didn't spec any of the equipment to be used. I did ask the question "can we use the spring pot we are using on the other system we have now as well as the load cell. I was told YES, no problem. The 13'000 over run has been billed to us. We did not have any input on them spending the time before it was spent. In the P.O. to them is states that any work/time over the original P.O. needed written approval from both sides. I agree that if they had came to us midway and told us they where having a problem with the string pot or ML1500 we could have worked those things out and known about the extra charge to do so. I had an hydraulic engineering firm work on the pump and they had there work done before the electrical side was done. That way everyone was on the same page. If the electrical would have had a problem with the pump side we would have changed that before anything was started. Everything was in written form before the work started. I really don't think its on the hydraulic side of things. I could be wrong. I am posting those specs. It was the Hydraulic outfit that recommended the electrical engineering firm to us. We have not used them before. the Electrical company that is. I made every attempt to make sure everything was agreed upon before anything started, with both of them. No P.O.'s where issued until everyone was happy. The hydraulic side knew what the speed needed to be and how we wanted to control it. The electric side knew what the hydraulic side was going to use and what would be expected of there controls. Or at the least I thought so. I would post the specs that where agreed to, but I have no way to remove the name of the firm in question and I don't think it would be right to do so with out removing there name. I would e-mail to someone if needed. Thanks everyone for your input it helps. Large_Pump___cylinder_J11718H_Model.pdf Edited by mracer

Share this post


Link to post
Share on other sites
Why not post ALL the data I was asking for. Don't bitch to me about the money. I/we can't do anything about it. If you won't supply the information then fine. I have other things to do. I think the programmers/electrical guys are ignorant victims in this too. You were all screwed by the hydraulic design. I will know more when you post the rest of the data. Since you are the customer you are at least allowed the privilege of being ignorant. That is why you hire so called 'experts'. In this case I don't think the experts did their job. It is the primarily the hydraulic design. There isn't an accumulator for starters. There is a regeneration circuit for second. If system must accelerate and decelerate quickly it doesn't have a chance. 23 (Gal/Min) x (Min/60 Sec) * (231 cu in/Gal)=88 cu in/sec. Unless this system has a very small diameter rod it will not move at 120 inches per second. If the diameter is small then there will not be enough area to provide enough force to accelerate under control. You haven't provide me with all the data I asked for. No natural frequency. No performance specifications. No mass. No cylinder and rod diameter. The cylinder part number isn't good enough unless you can send me a link. I am familiar with the valve. The part number says is doesn't have electronic feed back. NO INNER LOOP for spool control!!! At high flows the flow forces will move the spool around. That is not good if you want precise moves. YOU CAN'T BLAME THE ELECTRICAL GUYS IF THE SYSTEM WILL NOT GO 120 INCHES/SECOND!!!! ALL THEY CAN DO IS OUTPUT 10 VOLTS TO THE VALVE. THEY CAN'T OUTPUT 11 VOLTS AND MAKE IT GO FASTER.

Share this post


Link to post
Share on other sites
Off topic, but wanted to ask since you mentioned it Peter. What is "natural frequency". I am pretty sure I understand the basics of the other things you mentioned and can see the effects they would have on the system. But am totaly lost on the natural frequency thing.

Share this post


Link to post
Share on other sites
Peter, I respect your input. Thank you. I hope I haven't came across as wining about the cost. If I have I'm sorry. My job now if to figure out what to do with a project that has gone wrong. If that means a differant valve, thats what we will need to do. The problem is I really dont' understand who to trust now. I'm not an engineer. I am just a maintenance guy that is learning PLC's and HMI's. I thought by going to engineering firms I would get was asked for. I'll post what was agreed to. If posting this upsets soemone , well so be it. If it will help others to understand where I am at. I guess I'll take the heat for that when the time cames around. Microsoft_Word___07_20053_EST_068___Wenger___Hydraulic_Controls_Description.pdf Edited by mracer

Share this post


Link to post
Share on other sites
The natural frequency of an object is the frequency that an object natrually will resonate or oscillate. Hydraulic cylinders will oscillate. The oil on either side of the piston will compress and expand. The oil acts like a spring. Think of how you would control the other end of a slinky as opposed to the other end of a steel rod. Controlling the other end of a slinky is hard. Controlling the other end of a steel rod is easy because the rod doesn't vibrate. The goal is to design the hydraulic system so it has a natural frequency that is much higher than the frequency of acceleration. Things that make the natural frequency high are Using large diameter cylinders. Reducing the load mass. Reducing the length of the cylinder. Not usually practical. Keeping air out of the oil Things that make the natural frequency low. Air in the oil small diameter cylinders long cylinders. valve mounted far away from the cylinder. rubber hose between the valve and the cylinder.

Share this post


Link to post
Share on other sites
You are definitely not going to be able to hit a 120"/sec. An accumulator MIGHT help but a tremendous amount of additional details are needed It is going to take well over 6 seconds for the cylinder to move 120" while extending I would be very interested in learning a more scientific approach to troubleshooting this. Do you have any pointers or links to any good reading?

Share this post


Link to post
Share on other sites
http://www.deltamotion.com/company/news/pr_20070228.php The design guide is still very basic. Detail analysis requires solving a system of non-linear differential equations. I will be giving a presentation on hydraulic modeling and those non-linear differential equations at the IFPE show at Las Vegas on Wednesday.

Share this post


Link to post
Share on other sites
OK, I F-ed up on this one. It needs to move @ 36"/sec. I really don't understand why I am taking so much heat on this one. Now I even have Peter hot over it. I am giving as much information as I can when I can. If I am being a little slow with getting the data out there it's because I am still gathering it for all of you. And some I just do not have. Maybe I shouldn't have asked.

Share this post


Link to post
Share on other sites
Mracer - One big thing that may help is to give us a basic overview of what this is. Not in the budget right now, but I still have a ton to learn before I could absorb one of your presentations. Thanks

Share this post


Link to post
Share on other sites
First off, unless there were design changes that you both agreed to, they can take their $13K and stuff it. Second, in this situation what you've said is that you don't know the engineering side of it and that's fine. You need to pay for experts to do that, and that's what you attempted to do. Your mistake was in splitting up the project and not putting performance requirements into the contract. If you put a performance requirement/gaurantee into the contract for which the original PO was written, I wouldn't even pay them the original $17K. If you paid someone to build equipment to do X and that's what it says in the contract and it only does 10% of X, then they get zip, zero, zilch. Don't be surprised if they want to come and remove their hardware though unless you paid for that separately. That will at least extricate you from a contract gone bad. However, if you simply "paid for programming" or some such and didn't write a performance gaurantee, then consider it an expensive lesson well earned. I have 3 degrees: electrical, process (metallurgical), and business. You'd think that even though the only credential I don't have is a PhD that I'd "do it myself" a lot. There are three situations where I don't do that. First is when the job is "contained"...where I can easily write a spec to say that I need someone to do XYZ and where XYZ does not depend on a lot of other things that are out of the contractor's control. These are the very best situation all around. The second case is where I don't have the expertise in the first place. The third case is when even though I could just about do it myself, the time involved ties me up too much. In this case if I have a contractor that I trust to do it right, I'll do a T&M (time and materials) type contract with a not-to-exceed number (that I usually generate myself) and have the contractor do the project on that basis. Essentially I'm the GC in that case and any responsibility for the project going bad falls on me. In this case, I would have done a rough design just so that I can gauge what it's going to require. I would have written a spec (and got a PO) for a performance contract. In your situation, it should have been written as a turnkey contract. I would have even put those terms in the contract so that there is no question all around what you are expecting and paying for. That way you give them rough performance requirements and pay the money. You don't put any restrictions on implementation criteria and only expect a flat contact price. That way you have removed any responsibility for project failure from yourself. If it fails, you just tell the first company to go pound sand and then go to bidder #2. When you write up the project, use bidder #2's prices (or even #3) because it gives you some leeway if things don't work out, and because if you have to go to bidder #2, you won't be going back and requesting more money later. It is perfectly legitimate to suggest string pots or some such but in this case you've shifted all the responsibility onto the contractor. You've got to make it clear that it is only a suggestion. The reason being that for the speeds you are asking for, the string pot is probably WAY too slow. The particular application pretty much calls for a high resolution encoder with a high speed counter and not much else. In fact, the controller better be a motion controller too because you need very tight feedback to the control valves to be able to control a cylinder moving that fast. Even for an electric solution, depending on how accurately you want to control the stopping position, it might require a motion controller. Third, Peter is right. Unless the size of the load is very large, for the speeds you are asking for, a hydraulic system is probably a really bad idea. I would have personally gone the all-electric route. Even if you'd gone hydraulic over time as the valves and cylinders wear, you've got to design it for closer to 3 times the speeds/loads of the original design to allow for some mechanical degrading of the system over time. It should have been obvious that what you are asking for is very fast when you started or someone should have said something, but again, live and learn.

Share this post


Link to post
Share on other sites
I didn't say that using hydraulics is a bad idea. I just said it was designed wrong. The fixed displacement pump is too small to go at the original speed of 120 inches per second. It should be able to hanld 36 inches per second easily when an accumulator is added. The accumulator stores energy ( oil under pressure ) when the actuator is at rest. The energy stored in the accumulator can then supplemtnet the pump when the actuator is accelerating. This way the pump only needs to be sized for the average load plus a little more but not for the peak load like one would need to do with a servo motor. The accumulator has a second benefit. It keeps the pressure from fluctuating too much. A good design will keep the system pressure within about 10% of the steady state pressue. Keeping the pressure constant is important because the gain of the system varies with the pressure change. If the gain of the system changes then the PID gains should change too. This is not practical on a ML1500 and in any case an extra analog input is required so it is best to add a big accumlator. Accumulators will also absorb some shock. The current design is using fixed displacement pumps. When the actuator is stopped the oil is by passiing over a relief valve. This isn't very a efficient design. It is much better to store energy in the accumualtor. High end system may use a VFD to turn the fixed displacement pump that can be slowed down or stopped when the system stops. Even at the lower speeds of 36 inches per second an accumulator is necessary. The lack of accumulator was the first thing I noticed. This alone makes me question the rest of the hydraulic design. I am waiting for more details and facts before I make further comments. Moving a few tons 120 inches at 36 inches a second isn't too hard if the acclerations and decelerations aren't too high. Edited by Peter Nachtwey

Share this post


Link to post
Share on other sites
Ok, This part I understand. We have several Husky Injection Mold machines and we use ALOT of accumlators. Since we are doing thin walled products we have to inject very fast. We have added accumlators to several machines to assist in this. the way I understand accumlators is they supply X volume of oil at X pressure. I know the results are pretty dramatic. As always Peter, thanks for the lesson. I believe I understand what you are talking about now in reference to natural frequency. The last time some one eplained it to me they called it resonate frequency. It was my understanding then that load had the greatest effect. If sizing a cylinder you would want the cylinder to be able to generate force several times greater than the force of the load. I put in a request for the guide. Thanks for the pointer.

Share this post


Link to post
Share on other sites
Since your interest is real ftp://ftp.deltacompsys.com/public/PDF/Mat...20Frequency.pdf I put this on RSDoran's site a few years ago. http://www.patchn.com/downloads/natfreq.xls I have a lot more.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0