|
Post by Deleted on Mar 17, 2015 11:02:05 GMT
Most people on here design things . I just wondered whether some of you would be interested in talking about the design process .
Most designers use something quite near to a random walk technique - plough in anywhere and hope for the best . Works surprisingly well sometimes but hardly an elegant way of doing things . In industrial design it's not very reliable either as many famous fiascos of product design will testify .
There are better ways .
I'm not going to write a tutorial on this subject - just introduce the topic and see where it goes .
------------------------------------------------------------------------------------------------------------------------
A little background theory to begin with :
There are two procedures for design - Deductive and Inductive . And two classes of design - Adaptive and Inventive .
In Deductive design a solution is postulated and then evaluated to see if it is viable . In Inductive design the ideal solution is put forward in abstract and then built up into a real design .
An Adaptive design uses mostly variations of existing designs . An Inventive design uses completely original thinking .
More than 95% of all design work is Deductive/Adaptive .
Inductive design is a bit hard to grasp at first and little used but it is applied in specialist fields like electronics and jet engines . It is by far the most sophisticated and effective procedure and needs be used more widely .
-------------------------------------------------------------------------------------------------------------------------
Carry on with this topic or drop it ?
|
|
|
Post by ejparrott on Mar 17, 2015 11:14:16 GMT
I'm interested, please continue!
|
|
pault
Elder Statesman
Posts: 1,496
|
Post by pault on Mar 17, 2015 12:08:57 GMT
Me to lets see where it leads us
|
|
|
Post by Deleted on Mar 17, 2015 13:31:33 GMT
Ditto! Let it continue! I have been using adaptive design method whilst retro fitting the vac system to the Michael Breeze V2 tender, it's good fun! Ben
|
|
|
Post by andyhigham on Mar 17, 2015 13:48:27 GMT
As Teflon Tony once said "there is a third way". I would call this the Microsoft approach to design, release an un-tested, bug ridden, poorly designed product to the market (windows 8), then release fixes to address customers problems.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 17, 2015 13:58:14 GMT
this might not be quite what michael had in mind, but the following link ive been following for sometime www.machineconcepts.co.uk/baldwin242/baldwin.htmit's the 'new build' 'LYN' for the Lynton and Barnstaple Railway in Devon. this isnt a pipe dream but is nearing completion. it incorporates some of the most advanced design and manufacturing processes. although in outline the loco will appear to be the same as the original 'LYN', the new loco has a gas producer boiler, piston valves, roller bearings, Lempor draughting, Porta wheel profiles etc. cheers, julian
|
|
smallbrother
Elder Statesman
Errors aplenty, progress slow, but progress nonetheless!
Posts: 2,268
|
Post by smallbrother on Mar 17, 2015 15:53:04 GMT
I used to "design" opencast coal mines. I was one of the few engineers (in this part of the world anyway) who was trained in just about every aspect of the process. It was a fabulous job.
I would like to think it came into the "inductive/inventive" category, certainly in the way I approached things.
As I learned however, you can be seen as someone who "over-engineers" things and takes too long. It is of course much easier to adopt what has happened before and chuck out a working solution in quick time. The "what might have beens" are easily ignored. The result was coal being left in the ground that is now beyond recovery. A bloody disgrace as far as I am concerned but the others furthered their careers much more than me!
|
|
|
Post by Deleted on Mar 18, 2015 10:19:32 GMT
As I learned however, you can be seen as someone who "over-engineers" things and takes too long. It is of course much easier to adopt what has happened before and chuck out a working solution in quick time. The "what might have beens" are easily ignored. The result was coal being left in the ground that is now beyond recovery. A bloody disgrace as far as I am concerned but the others furthered their careers much more than me! Sad but true . People who shout the loudest and most often always get the all the praise - doesn't matter in the least if what they say is pure twaddle ! Thanks to those few people that expressed interest in this thread but I think realistically I'll have to let it die now . MichaelW
|
|
|
Post by joanlluch on Mar 19, 2015 8:35:18 GMT
This is an interesting subject.
During my professional career I've been designing gas-liquid reactors for the precipitation (manufacture) or inorganic pigments. My general experience on chemical equipment matches the rate Michael suggests, 95% of all design is Deductive/Adaptative, or even higher. There seem to be some industries that reached some degree of maturity, and Inductive design is very rare. However, from time to time something that is clearly Adaptive on a particular industrial field can be regarded as Inventive in another field, though in reality it isn't. At a quite young age, I am proud of having put in practice such a thing: I managed to design a process that was not previously seen in industrial gas-liquid chemical reactors, but which was already at the experimental stage on biological wastewater treatment plants. Essentially I took an idea from a field and brought it to another one. But at the time, the reactor designs that came out from my idea were regarded as highly innovative (Inventive) and the same design brought me a lot of subsequent successes that basically marked all my remaining professional career.
I need to add that although I've been working in the chemical industry for decades, my speciality is computer programming. I mention this because I think that the fact that I was able to add an "invention" to a particular chemical reactor design, was in part because chemical processing was alien to me, and thus this allowed me to think in a fresh way that was probably hard for those who were in the field for years.
As per the model locomotive I am designing. I think that there are at least a couple of "inventive" aspects in it:
- One is the action (or mechanism) of the front bogie: It uses two spherical pivot points to allow both sideways movement and vertical rotation, avoiding all the classic design elements in there. - Two is the mechanism of the pony truck: Instead of using a single pivot point and reach a compromise on the place where it can be located, it uses two pivots to create a virtual pivot just at the place it should be.
As smallbrother already mentioned, and Michael corroborated, innovative ways to do things often get dismissed or brought to the dust by others, louder speaking ones. In my early labour days I was lucky enough to be given the opportunity to prove my idea which fortunately worked fairly well from day one with not a huge amount of tuning required.
|
|
|
Post by runner42 on Mar 20, 2015 8:12:19 GMT
I was hesitant in providing an input because the topic is something that I know little about. But if I consider what many of us do is build model steam locomotives and since all steam engine technology has been produced and documented, we are either copying or adapting existing designs. So at best we apply the adaptive design process and I shall be bold enough to say that inventive design process could only be attributed to the design engineers of yesteryear operating in the steam age. Unless there is some un-thought of device applicable to a steam locomotive that has yet to be designed. I don't know if using alternative modern day materials or say an innovative manufacturing process such as additive manufacturing would constitute an inventive design process.
It is a matter of history whether these design engineers applied a deductive or inductive process, as indicated in the opening post most applied the deductive process.
Does anyone know of the use of an inductive design process being applied to any part of a steam locomotive?
Brian
|
|
|
Post by joanlluch on Mar 20, 2015 8:38:12 GMT
Hi Brian, what you state is quite true.
Maybe one aspect where inductive design could have been applied on steam locomotives was on valve timing and events. We can think on a theoretically perfect set of events to open/close steam flow to the cylinders based on load, speed, actual steam pressure and more. Then we should be able to design a mechanism, possibly involving electronics and computer software, to actually produce such "perfect" events based on the opening of electrically (or other independent means) actuated valves.
Still, even if the above would probably be considered "inductive", it may not be considered "Inventive" as this is about what it is already applied to high end cars in order to get more power from the engine.
|
|
|
Post by steamcoal on Mar 20, 2015 9:28:20 GMT
I have spent twenty years thinking about designing a new factory for our manufacturing plant. Eventually got to build it two years ago. We had three machines and I placed them on the ground floor and built what was required well, from the ground up around them. It worked and there is really nothing I would change, except that I need a bigger factory building now!
It's not as complex as an oil refinery but it was satisfying to design it and finally work in the building.
Joan, I wonder how a PLC controlled steam engine would go…or how long it would go before it went AWOL? Anyone got an OMRON PLC lying around in a box? Not sure if that is inventive or just a nightmare.
|
|
|
Post by joanlluch on Mar 20, 2015 10:39:59 GMT
Yes, I think a small Omron PLC would do it with no issues.
If I was to do it I would do this:
1- Add a rotary encoder to the main axle, or alternatively linear ones on the piston rods. This would give the PLC information on piston position and speed.
2- Have a potentiometer to specify desired cut-off
3- Remove completelly the valve gear and the piston valve.
4- Fit fast actuating solenoid valves at each end of the cylinders.
5- Write software in the PLC to actuate the solenoid valves in real time based on piston/wheel angular position, speed, and cutoff (potentiometer)
That would be a basic design that would still use clasic arrangement of cylinders with connecting rod and coupling rods to the axles
The above could be enhanced by adding the regulator valve into the equation, actually driven by a continuous actuator, and taking into account the actual steam pressure as it reaches the cylinders
|
|
jem
Elder Statesman
Posts: 1,062
|
Post by jem on Mar 20, 2015 17:14:48 GMT
Steamcoal your factory sounds most interesting, would you be willing to tell us a bit more about it please.
Jem
|
|
|
Post by Deleted on Mar 20, 2015 22:06:42 GMT
We've had several response postings now so perhaps this thread should continue as originally intended .
Re: Steam locomotives .
Steam locomotives are almost(a) unique in the engineering world in that all the essential features of successful design were embodied in some of the very earliest prototypes .
There is no fundamental difference between Stephensons Rocket and Evening Star . The much later engine is bigger and a bit more complicated and a bit more efficient but otherwise the two engines are identical in design .
So the answer to the question " Does anyone know of the use of an inductive design process being applied to any part of a steam locomotive? " is essentially no I don't .
The only man ever to come close was Porta in relatively recent times . He started by determining what the ideal steam cycle should be for a practical engine and then designed engines to match with every minute detail optimised .
The great designers of steam locomotives had a depth of engineering knowledge which they applied very effectively but very few of them actually invented anything completely new .
note(a) : Jet engines are one of very few other instances .
---- More to come----
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 20, 2015 22:21:41 GMT
i would suggest that there are lots of examples of inductive design in locos.
GWR The Great Bear's boiler is one example as is Bulleid's chain driven valve gear for his pacifics, plus his Leader design.
whereas the process of development of the standard Churchward boilers was 'adaptive', the Bear's boiler was not part of this process.
there have been quite a few recent developments. i have been very privileged to have got to know Jos Koopmans re loco draughting, and he has completely overturned the traditional thinking on how draughting actually works in practice, plus found out that miniature locos actually lead the way in this respect because the margin for error is far less than in fullsize due to a number of factors.
as a bit of an aside, i always think that a firm understanding of cube laws etc and how velocities apply to miniature require far more attention than has often been the case. Jim Ewins had something to say on both, but i have a feeling Michael could say a lot more!
cheers, julian
|
|
|
Post by RGR 60130 on Mar 20, 2015 22:33:53 GMT
There is no fundamental difference between Stephensons Rocket and Evening Star . The much later engine is bigger and a bit more complicated and a bit more efficient but otherwise the two engines are identical in design . In one way Stephensons Rocket was more advanced than Evening Star because it had down-comers from the barrel of the boiler to the bottom of the fire box which gave a defined flow path as seen in the likes of marine and stationary boilers. In a normal locomotive type boiler all 4 sides of the fire box are (naturally!) hot and trying to make the surrounding water rise. In practice the down-flow probably occurs down the throat plate (weren't some throat plates left unlagged?). I have always fancied trying to do what in effect would be applying the science of a marine external superheat D type boiler to a locomotive outline boiler by way of using the front part of the barrel as an economiser then having external down-comers bringing the water down to the foundation ring. Maybe it would work better, maybe it wouldn't. I'll wait for someone to try it and tell me. Reg
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Mar 20, 2015 22:45:42 GMT
hi reg,
what you propose sounds very sensible and i can see the advantages so far as water circulation is concerned.
a bit off topic, but a sloping throatplate would achieve the same effect. Churchward did a lot of testing of these things and remained 'standard' till the Castle boiler due to weight restrictions.
when a Bulleid MN was tested at Rugby, they also tested a Bulleid MN without thermic syphons, and no difference could be found between the two types of boiler. if all Bulleid pacifics in preservation had no thermic syphons this would save an enormous amount of money in boiler repair costs!
cheers, julian
|
|
|
Post by Deleted on Mar 20, 2015 22:58:43 GMT
" Does anyone know of the use of an inductive design process being applied to any part of a steam locomotive? " is essentially no I don't . Hi Michael.. i was wondering... would Gresley's Hush Hush design not come under inductive design? or at least it's boiler or would it still be deductive due to it in part using marine boiler ideas? regards Pete
|
|
|
Post by Deleted on Mar 20, 2015 23:21:21 GMT
I'll get back to you all on points raised in next day or two .
MichaelW
|
|