WSF/FUD 3D printing - Please Ask don't grumble!

[JCL]

Hi Simon

 

This has been a really interesting, and given the topic, a very good natured read.

 

Although I'm not doing anything at the moment, I've put some models out there in the past for people to buy, and it was one of my models that was the subject of Dave's (Chris P Bacon) resizing point. I created a bodyshell and chassis for a GNR B3 2-2-2 locomotive. Dave and I bought our models at the same time using the same uploaded file, and while Dave's was absolutely fine, i.e. the right size, the bodyshell on mine was, while beautiful in its own right, about 3mm too short - way out of tolerances and the chassis didn’t fit. They weren’t done on the same machine because my print came from New York, Dave’s from the Netherlands. To get a reprint, all I did was send them a photo of the model against a ruler and told them what it should have been.

 

Other Simon, Shapeways have a satisfaction guarantee, so if you aren't happy with your print, then you ask for a reprint/refund. They do keep tabs though to make sure this isn’t abused I’ve found Shapeways support very good, even if their output is variable.

 

OP Simon and I have different ways of working, and I’m amazed at his output. I haven’t designs 10% of models Simon has out there! Also, I still only design models that I want for myself.

 

With regards to testing before making models available, except in two cases, I have been able to print everything I’ve modelled before I put it on the site for sale. For the things that I produce, I just about make enough money from previous models to pay for the prototype print of the next one. Maybe you know different Simon, but even with your amazing output, I can't see how anyone can make a decent living from this. Given the amount of time it takes to put one of these things together, the hourly rate would be in tens of pence. Or maybe I’m just very slow! If this is the case, then these models are either a labour of love for the model, or for the process of designing the model.

 

The two cases where I didn't do a test print were a set of domes and chimneys for a Scottish railway company, which I was asked to do, and an Ivatt GNR large diameter boilered G1 0-4-4 locomotive. I had already produced and printed the small boilered version of the locomotive, so I knew that that version was good, and didn't personally need the large boilered one for my layout.

 

The thing that I did do, was to tell the potential purchaser that I was happy to make the variant in my spare time, but I would not be able to test it. He was happy with this, and the outcome was good. Similarly, if I made any of my models in a larger scale, then I wouldn’t be testing them – buyer beware in that case, but buyer beware with the full facts in front of them. To do this I’d either have a conversation with the buyer, or if I put them on general sale, I'd also put the text in red or similar at the top of the description, so they can't miss it. Yes, this means that the potential owner is “doing my dirty work”, but they would be doing this with their eyes wide open.

 

I don’t have a lot of spare time in my life, so I design once, and do it for FUD. If a design works in WSF with only a couple of amendments then great, I’ll put that up too, but if it requires a lot of rework then it doesn’t go up.

 

Also, things change, and models that did print ok can change to being unprintable because of new rules or a model checker having different views on what can be printed and what can't. I’ve recently had an email from Shapeways to say that a chassis that I know has been printed 4 times is now not suitable for printing. I don’t have the energy or time to argue, so for now I’ve removed it from sale until I get time to redo the bits that don't work for them now.

 

On the FUD/WSF "which is better?" conversation, for me it’s horses for courses, and to be honest its not worth the worry. If your primary concern is cost, you’re happy to put in more work, don’t need the ultra-fine finish, and a WSF version is available, then buy in WSF, if the opposite is true, then buy FUD. No you won't get such a detailed model in WSF, but this isn't always the main motivation. Anyone that offers advice on this should first find out the primary motivations of the person looking to make a purchase.

 

Finally, for the designer/manufacturer discussion, I think we are missing the point. Instead of talking about role names, we should be talking about responsibilities. First of all, the designer is a business owner because he/she is having things made that people buy. The fact that they are ultimately made and sold by Shapeways is by the by. On top of that, the designer is responsible for getting the model right, and the manufacturer is responsible for getting the model printed exactly as the designer designed it. If the design is wrong, then it’s the designer’s fault. If the design is good, but the buyer still isn’t happy, then it’s worth asking yourself if the buyer was given enough of the right information for them to make the appropriate decision - were expectations managed. Because you aren’t a mind reader, it’s impossible to get this right every time, and of course some people just like to be triggered. (I think that's what they call it on t'Internet.)

 

Does any of that make sense? I'm not sure that it does, and if you managed to get through my ramblings this far, thank you for persevering!

 

cheers

 

Jason.

Edited March 13, 2018 by JCL

[rue_d_etropal]

Another dimension to the question of getting mdesigns correct, is scale drawings. Originals are not always correct, working diagrams are pretty close, and you just have to assume other drawings are as good as possible.

I am just doing some redesign on my Gazelle coaches, as Les Darbyshire has just updated the drawings he originally published in MRN. He still adds the rider that they areb ased on photos and the limited amount of info around, so they should be as close as they are going to get.

Some are critial of Skinley drawings, but at leas they exist and are probably still better that attempting to do a drawing purey based on a limited number of photos.

 

Another side of the ' please ask' idea, is that sometimes old original drawings, long forgotten about, turn up, as happened with the Ford Dagenham diesel at KESR. And this was after the superb model built and featured in BRM. On the other hand some supposedly originaldrawings are full, of of detailed errors, but end view is not symetrical, leaving one to wonder what tumblehome is correct.

 

I do feel responsible for my designs, so will not do something I am not comfortable with. Most modellers should be able to fix small details which are not quite correct. I build models to run, not to sit pretty on a shelf, so sometimes have to beef up some bits. What I do know is that if I do something one way, then someone will want it differently. If I don't put bufferes on, then someone will ask for them, and anyone wanting better buffers should be able to source them, but someone who just wants a model of something they can't get, then they are more than happy to accept the buffers as fitted.

[JCL]

Originals are not always correct...

 

You're telling me!

 

My method of working is to create boxes based on the dimensions given in drawings or from accompanying text, and then fill in from there. Often views that should be symmetrical aren't, and sometimes scanning a drawing might introduce stretching on one dimension. We have to take into account what the original drawing was intended for when basing models on them. I also use photos to fill in details missing on drawings. If I have a photo of the side of the vehicle then I'm in luck, but even with those we are fighting against perspective, so really we are looking at a "best guess" scenario.

 

 

Most modellers should be able to fix small details which are not quite correct.

 

 

 

This is true, but in general terms, if one finds out a model is incorrect, then morally, it should be taken off sale and corrected.

 

On buffers, I include holes in the buffer beams and a set of dummy buffers on a sprue - the sprue being the same diameter as the hole and 3rd part butffers, then they get the best of both worlds. With regards to lots of changes, then sometimes you just have to say no.

Edited March 14, 2018 by JCL

[sem34090]

The only reason my soon-to-be released G6 has the equivalent of 'moulded' buffers is because making a hole in the bufferbeam was causing printability problems.

[rue_d_etropal]

I have a lot of 'hobby' experience,not just model making, but building layouts, exhibiting them, having articles in magazines and running a model railway shop. I have met, and talked to a wide range of modellers, and observe how people react.

I have noticed that often when you assume someone should be able to do smething, they can't, fail and give up. Help them half way, and it might not be perfect, but they continue and complete what they are doing, and more importanly don't give up. Setting your sights too high is a recipe for failure.

The most popular stand at one of our local shows used to be Crownline. People were happy to modify and fix problems with design .Now with so many super detailed t2r models, the market for detailing has declined, so it is more dfficult to get thos bits. Coupled with many local model shops closing over past 10-20 years, actually finding what you need is more difficult. Local model shops wr not just places to buy, but to get advise. Online forums can offer advise, but are often slanted.

As with any information/documentation, if you know what you are doing/know where you are going then you tend to not notice that that information/documentation is not as helpful to someone who actually needs it. Someone starting out in the hobby might wonder where they can buy buffers, and then finding suitable ones. Having buffers on a design is one less hurdle for the average modeller. The skilled modeller, who is possibly more concered, is not only capable, but is also in the minority of those in the hobby.

 

Finalled, a sad day today, as it was announced that Stephen Hawkin died yesterday.

One of his quotes I think is relevant

 

4. “One of the basic rules of the universe is that nothing is perfect. Perfection simply doesn't exist.....Without imperfection, neither you nor I would exist”

5. “People who boast about their I.Q. are losers.”

 

[stewartingram]

Just a thought if I may?

Having just acquired your Wickham 109 cars (thanks very much for filling my gap), it has occurred to me that it might perhaps be better to produce a "kit" rather than a one-piece body? I'm no expert in 3d, but I believe the cost of the print is dependant on volume. Therefore 2x sides, 2x ends, 1x roof must surely be less volume than a full body? Also, going back to DC Kits dmu kits, the sides were produced as short panels, partly due to limitations of size of mouldings I believe.

Don't take this as a criticism of what I received, I happy with what I've got. Any minor errors that I find can I'm sure be corrected. I'm only thinking of possible savings on cost with future models.

 

Stewart

[Annie]

I was delighted to read Simon that you are working on a model of 'Gazelle'.

[sem34090]

Saves me finishing my one!

[chris p bacon]

Just a thought if I may?

Having just acquired your Wickham 109 cars (thanks very much for filling my gap), it has occurred to me that it might perhaps be better to produce a "kit" rather than a one-piece body? I'm no expert in 3d, but I believe the cost of the print is dependant on volume. Therefore 2x sides, 2x ends, 1x roof must surely be less volume than a full body? Also, going back to DC Kits dmu kits, the sides were produced as short panels, partly due to limitations of size of mouldings I believe.

Don't take this as a criticism of what I received, I happy with what I've got. Any minor errors that I find can I'm sure be corrected. I'm only thinking of possible savings on cost with future models.

 

Stewart

 

 

If you've had the DMU printed in WSF that would not be a viable option as you'd be unable to glue the parts together effectively. Mike Trice has produced some GN 6 wheel coach parts in FUD in the way you describe, this is so they take advantage of the better detail that can be printed.

[JCL]

I was delighted to read Simon that you are working on a model of 'Gazelle'.

 

It's an odd little thing isn't it? I always imagine Noddy and Big Ears riding on it.

[Annie]

Oh yes indeed  

 

Now I can't get that image out of my mind.

[rue_d_etropal]

I was delighted to read Simon that you are working on a model of 'Gazelle'.

Gazelle is one loco I have wanted to model for over 30 years. I even enlarged the MRN drawings to 1/19th scale as an idea to work with my SM32. The drawings for the loco in MRN are OK, but the original coach has errors, now corrected by Les Darbyshire and can be seen on the Colonel Stephens website. The other coach has some debatable errors, as different people have measured the coach over the years. I have been talking to one of the CS members who has told me quite a bit. In reality it is only a difference of an inch or so, so makesno difference in a model. For me, I was also looking to do the Wolseley railcar(same body), and the published drawings are showing a different length. I am sticking with the ones I have for when it was a coach.

It does seem you have to be very careful with any published drawings. Not always easy to compare to photos. The Drewry railcar trailer from the WCPR is another good example. Drawings dne in 1951 are way out, but I think between me and CS society I think what I have done is close enough.

Needless to say I have an application form for the CS society ready to fill in and send off.

 

I have said it before, but 3D printing is best for complete bodies(maybe a few bits to add), rather than a kit. Although there do seem to be people building kits, there are definitely fewer these days(according to one well known manufacturer/trader I was talking to last year). Where 3D printng has a problem price wise is OO, whereas N is small enough to be cheap enough, and O/G1 are big enough to not waste space, assuming you design with this in mind. Of course it is also possible to produce models in any scale inbetween. I have done one request for 1/48 th scale, and if anyone asks will see if others can be resized. For the hobby this means you are not limited by a small range of scales.

For anyone interested, check out latest Continental Modeller, where there is an article about bulding a Cramton loco, from a 3D printed kit.  I say , kit, but the body is one piece. It is HO, but in theory it could be easily upped to 4mm/ft .

It is still early days for 3D printing and scale models. It won't be long before  it can replace injection moulded and resin manufacture.

[chris p bacon]

It is still early days for 3D printing and scale models. It won't be long before  it can replace injection moulded and resin manufacture.

 

 

I doubt it will ever do that, the low production costs of injection moulding for large production runs and resin for medium are hard to compete with.

[rue_d_etropal]

I suggest anyone who wants to set up a business using injection moulding for models actually does so, not expect someone else to do so. Look at the problems some are having, and the number of retirees over past few years.

If you can produce models to order, not in batches which sit on shelves getting dusty, then injection moulding is sustainable, but the high initial cost of set up would not make that practical. That's why we have had white metal and resin casting for years.

Assuming you have created your moulds for injection moulding, released the model, and someone spots an error. Do you fix it, and at what cost? This actually has happened and it has not been cimmecially viable to fix it. With 3D printing to order, a design can be fixed, and all subsequent models use that new design. Cost of that fix is minimal.

 

HP do have a 3D print machine that can print to high quality, and Shapeways have been testing it.

 

Developement tie is another issue. Some might be able to create a master for resin casting quickly, but for injectuion moulding it takes a bit longer. Then there is production, and look at how long it takes our r2r model companies to design, produce, and then ship products. Sadly there is a trend these days for people to want things now, and not have to wait.

[AndyID]

...and you can also do some things with printing that are almost impossible with injection molding.

[chris p bacon]

 

Assuming you have created your moulds for injection moulding, released the model, and someone spots an error. Do you fix it, and at what cost? This actually has happened and it has not been cimmecially viable to fix it. With 3D printing to order, a design can be fixed, and all subsequent models use that new design. Cost of that fix is minimal.

 

The inference with that though is that less care and research is undertaken as the design can be adjusted for subsequent prints.

 

That is more likely to leave a sour taste for the purchaser and bad news spreads quicker than good. 

 

This goes back to test printing your product.

[Knuckles]

I've just read in one blast both of these pages with interest so as I am involved I thought I may as well throw my view out also.

 

Regarding test prints I mostly do but not always. All the reasons stated by everyone on here are my own views also, you may notice I hit the agree button a bunch of times.

 

It is all case sensitive regarding what is test printed. When I first started off I test printed a lot and in the process learned oodles and spent a lot on the process.

 

Now I can make things and not need to test them as much but I still do the majority as I want most things for myself.

 

I was asked to rescal the extended tank E2 to N and found the backhead detail really needed eliminating for motor space so I took a mini drill to it and made the space. Then redesigned the CAD to copy what I just did in reality then re-uploaded it. Didn't need a test print at this point so didn't order one.

 

If you look at my Furness K2 / 21 Class body you will notice it has 3 real versions; the standard one, the 'late' one and the Phoneix Superheated one with the same boiler pitch. A higher pitched type may come later. On top of that there are 5 Freelance Variants I made totalling an extra 10 meshes as I optimize certain things for both WSF and FUD/FXD prints.

 

As the body geometry underneath is universal for my own chassis there is ZERO need to test print every body when the detail variations are just that. Variations on aesthetics.

 

I also along the way saved up and purchased a Robox Dual (RBX 02) 3D printer and love it much. I use it mostly for it's original purpose in prototyping. My range of A0 Pacific locomotives were designed to fit several RTR A1 chassis and I printed dummy chassis and running plates on the Robox to test fit and tweak the design. I have yet to order an A0 from my own shop but I know things fit because my own print outs and at least two customers have already made very nice jobs of them.

 

So it depends really.

 

If I test printed literally everything then I might as well give up because we would be in debt. It isn't like we are making a killing on Shapeways as they hog a lot of the money for themselves most of the time. Ghat said in time as your ranges increase it may also increase the chances of making sales although this is also offset by the novelty of a certain design wearing off once everyone has brought one meaning you habe to keep the ball rolling with new products! Fun though.

 

What seriously grips my poowp though is people asking for re-scales or variations that I agree to and then decide they don't want to buy one any more. It has happened many many times now and the hours wasted is not funny so I'm more choosey and strict about things than I used to be. Most of the re-scales I currently am doing are speculative.

 

I'm doing a lot of them in hope. Fishing basically! If I don't get a catch no harm done but when I'm specifically asked or commissioned in some cases to do something and then they change their mind it isn't acceptable. For commissions and bigger jobs I'll be doing a 50% before and after jobby from now on. Was hoping to avoid being so anal about things but there you go.

 

On to materials:

 

I offer nearly everything in WSF & FUD/FXD as default, often with separate uploads for variations based on tolerance and detail differences. I do this as my SOP.

 

When going down to N though it is often a lot more work requiring thicker walls and less details and sometimes a significant rebuild in areas. I won't even attempt offering them in WSF, based on the amount of rejected WSF prints and customer cancellations (not the customer, Shapeways cancelling it at a whim) Shapeways thrown at us for the most dumbest of things, usually things that printed time and time again I son't even consider WSF for N. FUD cancellations/rejections are few thankfully.

 

I'd LOVE to offer my loco bodies in BHDA but based on my previous attempts (see the HDA thread I did) I don't know if it is worth it. I'd like advice and opinion on those who took it further with success if you are willing to give it. Quarryscapes comes to mind!

 

HDA seems a lovely material that is uber smooth and high detail. Better than FXD in areas but the sprue issues, rough clipping residue from supports and often (based on what I experienced and seen since) banana warping seems a bit of an off put.

 

I very much want BHDA to work for my range as I like the durability and forgiving nature of WSF (Great for beginners too) but also like the crisp smoother detail.of FUD/FXD on a good print. BHDA seems a good mix despite the issues.

 

Yeah that'll do.

Edited March 18, 2018 by Knuckles

[rue_d_etropal]

In many ways coaches are a lot easier to design than locos, once you have a rough plan(not scale, but sequence of design instructions). Most coaches are a long thin box, with a curved top. Sometimes roof is more complex.Then there are doors and windows, sometimes in end(s). a flat base and sometimes a basic chassis. Once I know that plan works I can be pretty certain it will always work. I know how much adjustment to make for different scales. For something more complex, I will get a print myself, but no need most of the time.

I am always happy to have a third opinion, and when someone has suggested a particular design, or has specific interest/knowledge of a prototype, then I will share scale drawings based on my design if they ask.

[JCL]

Hey Knuckles, I know we've spoken about much of this in the past, so I agree with everything you have to say. I know it's a pain to have to ask for money up front, but it does sort out the wheat from the chaff. 

[Gareth Collier]

Quick comparison between FUD & WSF in 'N' of the same coach. No remedial work carried out, just given a dusting of primer.

.

Edited March 24, 2018 by Gareth Collier

[rue_d_etropal]

I would not prime the WSF , certainly not before cleaning up. Even a bit of damp, can make it more difficult to remove loose nylon dust. First thing I always do, is add a drop of cheap superglue to buffers, as these can easily get damaged.The slightly narrower part behind buffer head, acts as a fracture line, and it needs strengthening. Learned from experience. 

[Gareth Collier]

I just gave it a dusting of primer as no amount of cleaning up was going to save it For about £4 there's no comparison. The FUD one is excellent, many thanks for making it available.

[rue_d_etropal]

I am looking at producing some finer detailed models for FUD and FED. Nothing in a hurry though.

I have found I can clean up WSF, and I have seen models others have done, and they look OK. Saying that,it takes only a few seconds for me to add FUD to options, and for N scale the price difference is not much. For OO the price doubles , soIwill only offer that if asked, and possbly for some models with very fine detail.

[AndyID]

For OO the price doubles

 

Considering the enclosed volume is about eight time greater, that's not all that bad. Although I know it's likely to be a major pain, one way to reduce the cost of FUD might be to reduce the height of the print by producing panels that have to be glued together.

 

Shapeways applies a significant charge for the height of a FUD print. It's proportional to the height raised to the power 1.25. That might not sound like a lot but it has quite a big impact on the cost of a tall model compared to the cost of a low profile model with the same volume of material.

[rue_d_etropal]

 

one way to reduce the cost of FUD might be to reduce the height of the print by producing panels that have to be glued together.

I think that is called a kit, not an almost r2r/aid to scratchbuilding model. Sorry that is not something I am prepared to do. The big advantage 3D printing has over kits is that complete(or almost complete) body shells can be produced. More in the hobby(and maybe those not already in the hobby) expect this. Look at the way good quality r2r has hit the kit market(I have talked to some in the trade).

FUD requires extra supports,and those cost money, not just to produce, but also to remove. The longer the support the more it costs.

Some might consider the extra cost worth while(not everyone is a competent kit builder), and compared to having to pay someone to build a kit, then worth the extra cost. I am starting to add FUD to options,partly because some might want it, and partly to show how much more it can cost. When someone asked for a FUD option on one of my big scale narrow gauge models, they were shocked by how much more it costed.

FUD has improved over the years since I started, and I remember a lot of grumbles about it, not dissimilar to what some are finding with other companies. Shapeways may not be the cheapest, but they reorganised a few years ago to ensure they could keep going, and that costs a bit more. Using the latest mahines(and testing protype ones), maintaining them properly all takes money and time. My only worry, would be if some big corporation thought it worth while taking over, and trying to cut costs, and quality. Seen it before.For me the business model that Shapeways use, and the way it enables designers to use actally works very well, and is sustainable.

 

A couple of years ago I suggested 3D print designers should get together and attend exhibitions. Unfortunately there was not much interest, but many of the problems,and myths about 3D printing could be discused face to face, rather than just online. I do show my 3D printed models at exhibitions, alongside and on my layouts. I think I might start including some form of demonstration of how to use the material, in partiular WSF and the ways I have developed myself. 3D printing is notjust about creating the base product, but how to use it. There are loads of demonstration stands with various modelling and painting demonstrations. I have never seen someone showing how to work with 3D printing.