Dapol Update News Shocker....

[adrianbs]

Hi Frobisher  I do fully understand what  a CAD is,  perhaps I did not make myself clear. If we the modellers are only shown  one CAD we cannot possibly be sure if is accurate dimensionally, if no dimensions are added or indeed, in detail,  if various different views are not shown. It is therefore difficult to comment on the likely accuracy of the model since we have only been given partial information..

  My last 2 sentences were, at least so I thougt, implying that the 3D modeller was more interested in producing an attractive image than worrying about the dimensions of that image hence the innaccuracies that finish up on the model when the tooling has been made which, as you say, should be the same as the CAD rather than that for an accurate model.  One particular problem that shows up on on finished models is the lack of variation in "Rivets" which appear on CADs with no apparent difference in identity whereas on ,the real thing there can be a vast difference in shape and size since they may well be small round headed screws or large hex or square nuts with washers and protruding thread. Since the CADs are often the result of combining different files and overlays there should really be some obvious differentiation but on, for example the Dapol POWs every single nut is exactly the same size and looks like a small round head rivet into the bargain rather than anything approaching the various sizes of hex or square nuts or even coach bolt heads which are more mushroom shaped..    Regards  all adrianbs

[frobisher]

  My last 2 sentences were, at least so I thougt, implying that the 3D modeller was more interested in producing an attractive image than worrying about the dimensions of that image hence the innaccuracies that finish up on the model when the tooling has been made which, as you say, should be the same as the CAD rather than that for an accurate model.

You really don't understand the process if you can honestly think that. These rendered images we get shown are a by product of the process, not the end point - the mere suggestion that the whole process is compromised to produce "pretty pictures" does a great disservice to the CAD operators as well.

[Jonboy]

Could the review of the R & D simply be because the new team are more comfortable with a different CAD software to Dave et al and so need redrawing? I could see issues deciphering someone else's notes in this situation?

[adrianbs]

Hi Frobisher, OK ! as these CAD images are of little use in determining the accuracy of the files used to produce them,, and as i fully agree they are a just an interesting use of those files why are they shown at all ?? Thats why I say it would be better if conventional end, side and plan views were shown so that we knew the starting point used to produce these "3 dimensional CAD images" .  There just seems to be total emphasis on posting multicoloured CAD images from which only a limited amount of information can be gleaned.    It is impossible to make an accurate 3D CAD view without all the basic dimensional information being available. Since many of the CAD images that have been posted clearly did not have the required information to make an accurate model, someone must be making assumptions or using innaccurate information which could be corrected if we knew what it was. If they are making assumptions it must be so that they can complete the various files and present something that will show off the planned design to the public ie a "pretty 3D CAD picture"  as well as producing the tooling.  As far as I am aware models such as the Dapol POWs, SR brakevan and Milk tank were not laser scanned for cost reasons.  Drawings or dimensions must have been obtained in order to produce the files unless of course, the whole thing is a figment of someone's fertile imagination, which in one case does seem likely. This would certainly have led to a "pretty CAD image" although in that particular case it has not been published so may never have been produced.

    I have worked with Martin Finney for many years and it is always the basic drawings for the parts which are prepared first, whether for patterns or etching artwork before any attempt is made to go 3D to check shapes and components fit together, all but the first model having been drawn on the computer.  Patterns are not made from 3D CAD images but from detailed drawings supplemented by photos although the drawings are done with a CAD program they could have been done manually or indeed patterns can be made with no drawings at all, provided all dimensions are known and photos are available. .Information is obtained in various ways from on site measurement to official GAs if obtainable.             Invention and assumption are a sure road to disaster.          Regads  adrianbs 

[Karhedron]

Thanks for posting. Do they say WHY so much R&D has had to start again from scratch?

 

Take the example of the N Gauge 33. Test-shots of the body have been shown and I do not recall any huge outcry or adverse comments, certainly it looked good to my admittedly limited knowledge of the subject. If I recall correctly (please correct me if memory is failing) the 33's chassis was on the verge of being signed off too? So what was so drastically wrong with the tooling that it had to be scrapped?

The following was posted on a different forum. I cannot vouch for its accuracy but I assume it is genuine.

 

Hi tooling error was to do with postion of exhaust - this a slide item for alternative versions so required slide moving -easier to get whole new tool made but since rest of data already held a "quick" change really just delay in getting tool making slot along with all the other work Dapol have shown as work in hand...

[Karhedron]

Could the review of the R & D simply be because the new team are more comfortable with a different CAD software to Dave et al and so need redrawing? I could see issues deciphering someone else's notes in this situation?

Certainly a possibility. I know from experience that 2 pieces of software that do the same thing may still do it in different ways.

 

There is also the "fresh pair of eyes" factor. If the above posting about the class 33 is true, it may indicate that Richard is going through the work currently in progress to make sure he is happy with it. In some cases this may require him to resarch prototypes he has not have any previous experience of modelling.

 

I am the first to go through any GWR prototypes with a fine-tooth comb but a 33 is a rare beast as far as I m concerned. If I was asked to approve a design for one, I would want to take my time going over the design and comparing it to the original as well.

[adrianbs]

Hi  Kathedron et al,  I am sure you are right about Richard having to get up to speed on projects he has never seen before. I think that is why he was trying to make it clear in the interview with Mark Chivers that things would not all suddenly appear totally redesigned and perfect within weeks. The problem Dapol presumably have is that they already had slots booked for different projects, and are not keen to cancel them as there may be a penalty and additionally they will not have new product to sell. Things have moved far slower for them than they originally planned with the departure of DJ and I am sure that cash flow is as important to them as most other concerns.  They have been left trying to progress a large number of different projects which unfortunately seem to need significant alteration with no time to do the work.

     In the case of the Hornby Magazine diesel brake tender it is clear that not only did the existing design need a major rethink but RW, a self confessed GWR enthusiast, found that there was inadequate information available from Mike Wild and he himself did not personally have immediate access to the required drawings which is why I was asked to help. He could, of course, have found a real bogie with no trouble and measured and photographed it to produce his own drawing,   I know how long that takes as that is what I did.  A copy of the official LNER GA does help speed things up enormously, especially when, as soon as you get back from a field trip you immediately find that you forgot to measure some vital dimension or take a picture of some complex component.

    It may well be that some of the designs are totally accurate and need little or no work but RW will not know this until he has gone through them very thoroughly using the available information. From what has been produced in the past he will be well aware that he cannot rely on existing designs being correct and he knows only too well that his reputation is at stake.  With the SR Brakevan, he has posted remarks to the effect that he had only been able to modify some aspects whilst other errors will still have to be accepted.  If only more information had been published earlier many of these errors would have been corrected already but as Dapol clearly had not made up their minds what they were actually doing and have totally changed the specifications of the SR B/van at the very last minute no time was left to do the corrections.  Having therefore chosen to do a version for which 3 kits already exist and, in all probabilty retaining the remaining errors they may find sales not as buoyant as they would like and this might make them rethink other variants which will cost a lot to tool up and may show diminishing returns.  Modellers really only needed one RTR SR brakevan type as long as it was available in a multitude of liveries especially if it had been a design not already produced. Slightly different variants with the same range of liveries  will only be likely to achieve a marginal increase in sales. Few people have ever pestered me to make the other variants planned, even though I could do so very cheaply because most of the tooling was designed to accomodate the LH ducket van.  In fact I only need to make a new cabin end and step support brackets.

         Regards all  adrianbs

[frobisher]

Could the review of the R & D simply be because the new team are more comfortable with a different CAD software to Dave et al and so need redrawing? I could see issues deciphering someone else's notes in this situation?

That's a possibility - also different factories may have different way of working with the resultant tools, and a different "stock" of common parts used within the CAD work.

[frobisher]

Hi Frobisher, OK ! as these CAD images are of little use in determining the accuracy of the files used to produce them,, and as i fully agree they are a just an interesting use of those files why are they shown at all ??

With respect, check these forums to see the feedback that has come from the membership over these images when presented. There has been plenty of feedback and questioning of the dimensioning and general arrangement and in quite a few cases, the issues raised were addressed.

 

Thats why I say it would be better if conventional end, side and plan views were shown so that we knew the starting point used to produce these "3 dimensional CAD images" .  There just seems to be total emphasis on posting multicoloured CAD images from which only a limited amount of information can be gleaned.

Perhaps they could/should present such views, but they are NOT the starting point of the 3D CAD work (they would be a set of views onto the 3D model). The starting point is likely to be the very types of diagrams you're talking about, but which the manufacturer does not hold the copyright for so could not reproduce those for all to see without permission. Producing functional duplicates of those diagrams from scratch to use as the basis to produce the 3D model from would just be an exercise in doing work for work's sake.

 

Any view, 2D or otherwise will only present a limited amount of information. A 3D images will show aspects of the shape, and general arrangement that are much harder to gleam from a set of plan views (Deltic's nose anyone..?). Would presenting a 3D model of the Heljan 47 have led to it's tubbiness being caught early on? Probably not as it does look like a 47. Would a set of plan views? Also probably not unless you also had a set of dimensioned plan views, and that's not something we all have in our possession.

 

These 3D views are not randomly coloured though (the colours may be arbitrary of course) - it shows the discrete mouldings etc. that will comprise the finished model, and it certainly gives insight into some manufacturing compromises that are being made.

 

I have worked with Martin Finney for many years and it is always the basic drawings for the parts which are prepared first, whether for patterns or etching artwork before any attempt is made to go 3D to check shapes and components fit together, all but the first model having been drawn on the computer.  Patterns are not made from 3D CAD images but from detailed drawings supplemented by photos although the drawings are done with a CAD program they could have been done manually or indeed patterns can be made with no drawings at all, provided all dimensions are known and photos are available. .Information is obtained in various ways from on site measurement to official GAs if obtainable.             Invention and assumption are a sure road to disaster.

You're talking about a very different process though. The common practice (certainly amongst the Chinese factories) these days is for injection moulding tooling to be cut directly by CNC machinery controlled using the prepared CAD work. Manual intervention comes in to fettle the resultant tooling where necessary.

 

But to reiterate - the purpose of the 3D CAD models is not to produce these images and any compromises and mistakes in the CAD and hence the resultant tooling and then model is not because they "have to" produce these 3D images. We are shown these because it is a convenient and relatively easy way of demonstrating progress in the production process to us the (hopefully) paying public.

[adrianbs]

Hi Frobisher,  I am fully aware that mass produced models are tooled using either direct CNC or indirect to make positives which can be pressed or spark eroded into dies.

 Martin Finney and I have found that this process is far too expensive for specialist models and the limitations of diecasting or injection moulding with inherent draft angle problems and the inability to have undercuts would reduce the quality of the product.   Whitemetal or Lost Wax or Resin  overcomes this problem to our satisfaction.

    I still hold to my view that many of the 3D CAD images are little more than pretty pictures provided to the general public to whet their appetites. This is especially true when little notice is taken, in spite of claims to the contrary, of crticism or queries of the accuracy.  The series of images of the Diesel brake tender clearly appear to show that it was designed without reference to accurate information and is therefore no more than a computer generated artists impression which in my parlance qualifies as a "pretty picture".  

        Regards  adrianbs

[Karhedron]

This is especially true when little notice is taken, in spite of claims to the contrary, of crticism or queries of the accuracy.  The series of images of the Diesel brake tender clearly appear to show that it was designed without reference to accurate information and is therefore no more than a computer generated artists impression which in my parlance qualifies as a "pretty picture".

I know that in the case of both the class 22 and the Western, a great deal of sttention was paid to the feedback generated by the CADs. In particular the shape of the cab roof on the western and associated air intake went through 2 or 3 iterations with Dave Jones spending a lot of time to make sure people here on RMWeb were satisfied.

 

I cannot comment on the brake tender as I did not follow the development. But to write off 3D modelling or to claim that the people producing them do not listen to input is certainly not true of the cases I have given feedback on.

[-missy-]

Hello

 

Just for information, the company I work for produces 3D models which are effectively treated as masters. Everything else is then subsequently produced from these models, be it 2D working drawings, tooling, etc, etc. The 3D is used as a very effective tool to get a 'feel' for the component, its size, proportions, and fit can be quickly judged and evaluated without the need for reverting to dimensional drawings. I am not saying 2D isnt used, dimensions are checked using a dimensioned 2D drawing but again, this is produced from the 3D model. 3D is by far more effective at checking clearances, tolerances, and fit than any dimensioned 2D drawing.

 

Julia

[adrianbs]

Hi  Missy,  I know you can do that, just as I produce patterns without any working drawings but are you saying that if you post a 3D  CAD image, of the type shown on these forums with No dimensions shown and no information about the scale that someone else can look at that image to determine if the whole of the model is going to be dimensionally accurate, the "rivet" detail is the correct size, and the definition of plank grooves and nuts etc is acceptable. I think not, although for sure if you send them the CAD file they would probably be able to backtrack to obtain all this information with no difficulty at all.   As far as that image on a forum is concerned it could be any scale and even with a programme to convert an image back to plan, elevation and end view  such a picture, for that is all it is, would need more information than has been provided. 

    If a single photograph with just one known dimension were able to be turned into a fully detailed working drawing or dimensionally accurate CAD file the inventor of that programme would be a millionaire within weeks.  Even the latest methods of producing scale models require a  number of accurately positioned photos and then can only show externally visible details. Even laser scanning needs much additional work to clean up the data.  All these pictures can tell you is if the planned model seems to be correct proportionally and if items are missing or should not be there,  ie whether it looks like a photo taken from the same angle if you happen to have one.  If it looks nothing like a  photo of the real thing but could be turned into a fully detailed model the model would be equally inaccurate and the image is no more than a pretty picture with no real value. 

   There were more than a twenty images of the brake tender posted but they were all obviously wrong and it was not even possible to work out what any of the major dimensions were but they were very colourful and would have looked quite nice framed !           Regards all adrianbs

[-missy-]

Hi Adrian.

 

Surely it depends upon what the manufacturer is trying to convey? If they are posting a 3D image to give people an idea of what the model is going to look like then it does the job. I think I take a different approach to things than you, I am not concerned about rivet size, plank gap spacings, or even bolt head sizes (but then I dont check use by dates on my food before I eat it or even check my hair colouring doesnt burn my skin before I apply it).  If it looks right to me then I am happy.

 

Julia.

[Les1952]

Is anyone here suggesting that we return to the days of Graham Farish (amongst others) where the first anyone knew of a new model was when a review of it appeared in the magazines, often failing to point out the boiler was over a foot too large in diameter or that bogies on a diesel or electric were sitting proud of the buffer beam?

 

Some of the posts are beginning to sound a little nostalgic.........

 

Just a thought.

Les

[adrianbs]

Hi All  Those modellers who have little interest in dimensions, details amd definition are well provided for by ACE Trains, Darstadt and Hornby- Bassett Lowke etc but there are many who would like something rather better and it seems unfair that we should be persecuted for this and expected to accept lower standards when there is no need.

      Regards  adrianbs

[pirouets]

Adrian.

 

If you don't want to feel persecuted then can i suggest staying away from post like your last one. As a modeller my knowledge on something's is extremely basic but my goal is to build a scene that pleases me. Hopefully I am better than I used to be but I don't care about some of the details because in n gauge I can't see it.

 

That is not to say I do not see the value that modellers with knowledge bring because since I came back to playing trains the standard of n gauge has gone up hugely and some of that does come from better engagement with modellers who have the knowledge as well as improved tooling and processes.

 

I do value seeing the CAD images to see what is coming but I don't expect manufacturers to post al the details. I do think Dave Jones set a high standard so the real question comes back to how will Dapol continue and for me, development pictures tell a good story of what is coming and I expect the manufacturer to get it right behind the scenes.

 

Also The question that does not seem to have been asked is IF Dapol would produce drawings with dimensions? Till that's answered isn't a lot of this thread kind of meaningless because we are guessing? Has anyone put that to the manufacturer?

[bigherb]

Can someone point me in the direction of the ABS models development thread.

[adrianbs]

Hi Bigherb,  Sorry! but now retired for 6 years and far too busy with other things to worry about new products. Can't keep pace with current range and can't work any harder as it would probably kill me.  Only post on these forums when I am too tired to do anything else. If there were anything worth watching on TV I would not even be on the forums.  You can blame the the BEEB or ITV if you don't like my posts,.   I only stir things on the forum in the hope it might result in better RTR models that I could buy and would consider of similar quality to kits etc.  Not had much luck in that direction so far, but I live in hope, especially if they do not duplicate models I already have with which I am well satisfied.

    Since most 7mm  RTR model manufacturers seem intent on duplicating existing high quality kits or other RTR manufacturers who have already duplicated these kits the supply of items worth considering is pitifully small.  If I were starting with a clean slate but working to the average standards of the better kits I would still find little to buy unless I was prepared to spend Mega bucks.   I could buy a fleet of diesels but there would be little RTR worth buying for them to haul or I could build a shunting plank with a couple of small locos and a dozen POWS from one range which has already seen many of it's products duplicated.  At the moment that is about it, In spite of all the RTR models that have been produced,  most of which are no longer in production.

    Rather than my development thread you might be more worried about how long any of my products will still be available, I shut down completely in 4 years even assuming I last that long !!!       Regards all  adrianbs

[Oldddudders]

Can someone point me in the direction of the ABS models development thread.

Ah, so that's who everyone but me has identified adrianbs as. I had a few ABS products, probably still do. They left me confused, because they appeared to be white-metal castings, not ABS (a particularly strong sort of plastic) at all.

[bbishop]

I think I had better beat Adrian to the keyboard, otherwise there may be a bit of a rant on the way. Remember that Adrian has been producing kits for about 50 years, so ABS Models probably precedes ABS the plastic.

 

Bill

[adrianbs]

Hi bbishop,   Only just !! adrianbs

[jjb1970]

These type of arguments tend to get polarised into extremes of the alleged "that'll do" brigade or "rivet counter" perfectionist when in fact few fall into either count. I don't know anybody of the "that'll do" brigade who does not want their models to be a good representation of the prototype. They may happily accept much lower levels of detail than others or accept compromises in terms of loco specific features, variants etc but they nevertheless expect a model of a train to look like the train it is claiming to represent. Equally I know of no "rivet counters" who refuse to recognise that making a model in OO or N means making some compromises to running qualities, details, manufacturing limitations etc. The truth is that most of us fall somewhere between the alleged extremes, further to one side of the line but nevertheless somewhere between extremes and it is a personal judgement of modelers where they stand on that line.  

[byron]

These type of arguments tend to get polarised into extremes of the alleged "that'll do" brigade or "rivet counter" perfectionist when in fact few fall into either count. I don't know anybody of the "that'll do" brigade who does not want their models to be a good representation of the prototype. They may happily accept much lower levels of detail than others or accept compromises in terms of loco specific features, variants etc but they nevertheless expect a model of a train to look like the train it is claiming to represent. Equally I know of no "rivet counters" who refuse to recognise that making a model in OO or N means making some compromises to running qualities, details, manufacturing limitations etc. The truth is that most of us fall somewhere between the alleged extremes, further to one side of the line but nevertheless somewhere between extremes and it is a personal judgement of modelers where they stand on that line.  

Generally I agree with your comments. However in the 21st century smooth running, with excellant electrical pick-up should be something we all should take for granted. Running quality is not an area for tollerance.

 

Mike

[jjb1970]

I agree on running. My comment was aimed more at things like axle spacing, wheel diameter, visual compromises to incorporate the motor and mechanism etc to allow for running. I like American models and some of the big US steamers (4-8-4 etc) have compromised axle spacing as true prototypical dimensions would be unworkable on a HO running model. There are examples of compromises to leading and trailing wheels on steamers so as to allow good running. And there are quite a few models that have had to make visual compromises to incorporate good mechanisms. However I agree that in terms of electrical pickup and smooth running this should be a given on a modern model.