What does "engineering accuracy" mean to you?

Discussion in 'Hydrodynamics and Aerodynamics' started by Leo Lazauskas, Feb 4, 2013.

  1. Petros
    Joined: Oct 2007
    Posts: 2,936
    Likes: 148, Points: 63, Legacy Rep: 1593
    Location: Arlington, WA-USA

    Petros Senior Member

    I find that often there is such obsession, especially with mono hull sailboats, to make each side exactly the same shape. A monohull, and to a lesser extent a cat, is never symmetrical in the water when underway except for going dead straight downwind. So it seems to me symmetry is way overrated.
     
  2. wardd
    Joined: Apr 2009
    Posts: 897
    Likes: 37, Points: 0, Legacy Rep: 442
    Location: usa

    wardd Senior Member

    to me it means nothing by itself, +/- .001" means something
     
  3. murdomack
    Joined: Jun 2007
    Posts: 309
    Likes: 23, Points: 18, Legacy Rep: 282
    Location: Glasgow

    murdomack New Member

    Times have changed. For many years now, a lot of large ships and structures are of modular construction. Obviously there has to be dimensional control, If part 1 is 1/8"over and part 2 is 1/8" under then there is a 1/4" mis-match.
    When I go round checking contractors fabrications and something is a little bit off, I don't want them to rework it if I can help it. I will go and find out what it matches up to, and if there is some give and take I will accept it. If it matches up to something that is fixed or another fabrication, I will make them put it right. Sometimes it hurts me more than them if there are time constraints.
     
  4. jehardiman
    Joined: Aug 2004
    Posts: 3,034
    Likes: 547, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Things really haven't changed, you still just dog it up and weld. Let's say we cnc plasma cut 2 18mm plates for the bottom shell of a module 12m wide +/- 5mm, @ 50C waterbath and crane them outside to the platten at 20C. You land them on the platten to get a proper root gap of 0-5mm. If you were to measure afterwards, the distance accross the 2 shell plates would be
    Minimum (24000-5-5-6-6-0-1)= 23.977m
    Maximum (24000+5+5-6-6+5-1)= 24.002m
    or a possible acceptable tolerance of 25mm about an inch.

    Of course, there would be a real measurement taken when we land the side shells, and the side to bottom weld would be trimmed or buttered to make up the proper distance and when the module was landed on the ways the two shells would be drawn up with dogs. In this way there really isn't any real difference between the old and new methods, it is still shipfitting to get critical measurements within tolerance
     
    1 person likes this.
  5. erik818
    Joined: Feb 2007
    Posts: 237
    Likes: 20, Points: 0, Legacy Rep: 310
    Location: Sweden

    erik818 Senior Member

    The part about accurate engineering I find the most difficult to accomplish is to take all relevant aspects into consideration. Calculating an answer to the required precision is seldom difficult once the problem is correctly set up. It's a dream to have engineered a complex system and never have to think "didn't think about that".
    Erik
     
  6. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 150, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Were there no high-precision tables of square roots available that would have sped up calculations?
     
  7. Ad Hoc
    Joined: Oct 2008
    Posts: 7,047
    Likes: 975, Points: 113, Legacy Rep: 2488
    Location: Japan

    Ad Hoc Naval Architect

    Which is also material related.
    Since a 25mm tolerance you've described would not be acceptable in aluminium :eek:
     
  8. jehardiman
    Joined: Aug 2004
    Posts: 3,034
    Likes: 547, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Hell, a large hull sags on the blocks more than that. When you land a block (say 20m x20m x 15m, ~300T) there are all sorts of things that are +/- 10-20mm. And I'll bet a 50m aluminium yacht is not +/- 12mm when all is said and done.
    I've taken measurement of ship in 1/1000 of an inch and I know how hard it is to get repeatable measurements. I know for a fact that if you measure the circularity of a hi-alloy submarine hull that the ship is an oval shape when on the blocks and more round when floating. You never truely level the block caps because of the curvature of the earth for things like CVNs, difference between bow and stern is ~2".

    Really, shipfitting vs machining is like carpentery vs cabinetry.

    So Leo, what were you after in this discussion? What tolerance gets built in, or what calculation tolerance is acceptable? They are two different things.
     
  9. Ad Hoc
    Joined: Oct 2008
    Posts: 7,047
    Likes: 975, Points: 113, Legacy Rep: 2488
    Location: Japan

    Ad Hoc Naval Architect

    It depends if your referring to overall dimensional tolerance of the whole hull, or plate alignment/gap tolerances. I was referring to the later :)
     
  10. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 150, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    I was moved to ask because I worked with a rowing shell builder who once
    told me that he didn't think carbon-fibre hulls could be made to the high
    accuracy he wanted. 17 metre long shells can sag by around 25mm
    at midships with a large crew on board, so his gripe amused me.

    I'm not after anything specific in this thread. I was just interested to hear
    people's views and maybe tales of disasters or unexpected serendipity.

    It also helps me decide on how many decimal places to produce in the output
    from some of my programs. I need to handle everything from 1 metre models to
    full-size ships.
     
    Last edited: Feb 6, 2013
  11. Petros
    Joined: Oct 2007
    Posts: 2,936
    Likes: 148, Points: 63, Legacy Rep: 1593
    Location: Arlington, WA-USA

    Petros Senior Member

    At large companies they develop internal policies on what is acceptable tolerance based on experience and practical fabrication limits. But sometimes this internal policy becomes more important that actually getting the parts to fit properly. Here is my story of "engineering accuracy".

    Years ago when I worked at Boeing designing retrofit repair kits for in-service aircraft, it was discovered on the delivered aircraft that the six foot long seals between the engine strut and the engine cowl did not fit properly and was getting damaged very quickly. These seals are in the supersonic blast of the turbine fan, and would allow the engine strut to "oil can" and cause all kinds of internal fatigue cracking, a very costly repair. the engine strut assembly was made of about 1200 sheet metal parts, bent up and riveted together. Is a major structural assembly since it holds the 55,000 lb thrust engine to the wing, plus it holds a lot of hydraulic and other equipment. I made a kit that required the operator to drill out and discard the holder for the seal (called the flare skirt fairing), and than install a new part that had no pilot holes. The mechanic would clamp up the new part with the correct seal clearance, and than match drill the fastener holes and than install the fasteners. this would give a perfect fit to the seal with the correct amount of compression.

    In my investigation I discovered the seal installation had a accuracy installation requirement of plus or minus 0.080", and yet the standard sheet metal part tolerance was 0.005". The chances of building up an assembly of about 1200 parts (each with a tolerance of .005) and get a final overall accuracy of .080 was almost zero. They were spending 80 man hours per engine strut in the factory reworking these parts to get them to fit, and most of them were still leaving the factory out of tolerance. Until the operators bought the repair kit I designed and installing according to my instructions.

    For some reason no one in the company thought having a tolerance of +/- .080" on a large built up sheet metal assembly unreasonable, I guess that was for "someone else's department". The strut was assembled in Winnipeg Canada, shipped to Renton, Washington where the equipment was installed, than shipped to Everett Washington where it was hung on the wing in final assembly. Than the last thing that happened was the engine was installed just before it was rolled out of the factory. Only than would anyone know if the seal fit properly, there was no way to measure it ahead of time, and there was no adjustment to the parts. So very few of these seals would never fit correctly, so it was costing the Boeing millions a year in rework, and the kit I designed was also sent out under warranty (costing Boeing even more).

    Hoping to save the company money, and save our customers a lot of frustration, I turned in a suggestion to NOT install the flair skirt fairing in Winnipeg, but rather install the part I designed (same part but without the pilot holes) with the engine kit, and simply align it before it was drilled to have the correct seal compression. There was no way a tolerance like that was going to be held on that big of an assembly, it just made more sense to adjust it to fit after the engine was installed. It would be the same number of parts and the same amount of labor (less all the rework), it was only a planning change on the assembly sequence. The company spent 4 years reviewing all of their manufacturing process, did a complete drawing and tooling review of all those parts; looking for a reason why their parts do not fit at the end of the process. Obviously, they assumed, there was something wrong with how they were making them rather than just change the assembly process as I suggested. After everything in the process was found to meet the company tolerance standards, they gave up on trying to find their "error", the agreed to the simple paper work and planning change I had suggested.

    Many times large companies do really stupid things. Once a bad process gets started, particularly across different divisions, it is almost impossible to fix the problem. they spent more money trying to fix the blame than to just get the parts to fit correctly by accepting my suggestion. There was four years worth of production that went out the door before the assembly sequence was changed.

    This is my story on what "engineering accuracy" means.
     
  12. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 150, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Splendid tale, Petros!
    Do you think a lot of the problem had to do with the fact that it was an
    aeronautical application, and hence the company had to go through a lot more soul-searching than other manufacturing industries would not?

    Back on land...
    I love the weird "Metropol Parasol" but it really makes me wonder what they
    had to go through before it was ultimately finished. That's a lot of wood and
    glue!
    http://en.wikipedia.org/wiki/Metropol_Parasol
     
  13. Petros
    Joined: Oct 2007
    Posts: 2,936
    Likes: 148, Points: 63, Legacy Rep: 1593
    Location: Arlington, WA-USA

    Petros Senior Member

    No, the fit of the seal was being delivered out of specification, and it caused some serious problems because of it. If there was a fix that could have been done by one department, it would have been done much sooner. It was the size of the company, and the number of departments involved that was the real hold up. A problem like that has to be solved by sending recommendations up to management, which would consider it, than send it back down the other side in the other department, who would form a committee to investigate their end of the problem and see if their department was at fault, if not they would send it back up the management chain where it would than makes it way done the organization in another department, and so fourth.

    what a company the size of Boeing (some 150,000 employees) does well, it does it because of its size, and what it does not do well, it is also because of its size. It can not respond very quickly to changes that disrupt procedures across multiple departments. A production problem that can be solved in one year is considered responding quickly. consider that it takes about 5 years from aircraft order to delivery, it can take up to a year to just get the delivery contract approved. the airlines are also large clumsy companies with almost as many employees. It is amazing they manage to make a profit and stay in business, and mange to keep everyone remarkably safe.
     
  14. murdomack
    Joined: Jun 2007
    Posts: 309
    Likes: 23, Points: 18, Legacy Rep: 282
    Location: Glasgow

    murdomack New Member

    I remember back in the sixties when I was an apprentice, they always referred to "Shipbuilding and Engineering" and based on what you have just said it must remain so.

    Most of what I deal with is structural or piping, so it will be well supported and accuracy is required as it will often have to be fitted using none flame work.
    I can understand what you say about sagging, but I am surprised that your procedures will allow you to dog hull plate if it is say EH36.

    Regarding accuracy, I worked in an oil fabrication yard in the nineties and we built and commissioned a complete deck for an oil platform that was going to Danish waters. The gravity structure that the platform was to sit on was built at another yard some 50 miles away. They surveyed the bottom of the deck and the top of the legs, and built all the adjoining pipe spools etc. We built load bearing access scaffolds below the deck and the finished spools were secured on these. The spools all fitted perfectly after the deck was placed on the legs, in fact only one weld had to be re-made offshore and that was on a pipe off one of the gas compressors that was pulling the allignment out of tollerance, probably caused by movement during the lift.

    Accuracy saves the clients millions of dollars and helps to improve the finished quality. I am surprised that the shipyards are not using grillages to support their pre-fabbed sections so that they can be moved around without sagging and distortion.
     
    1 person likes this.

  15. daiquiri
    Joined: May 2004
    Posts: 5,373
    Likes: 252, Points: 93, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

    Sounds like an overkill to me. What is the reason for such a tight accuracy on beam?
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.