Why do commercial production runs of <1000 models outsource to China / use injection moulding over 3D-printing?

[ShaneC]

Just trying to understand the logic behind producing the models in China / Asia using injection moulding vs. doing it locally using SLA printing.

It seems sort-of crazy to spend €20-50K (if blog posts from British producers are to be believed) for runs of 500-1000 locomotives on moulds alone and having 12-18 month lead times on new products. Is it to do with labor costs associated with the painting & assembly that more pushes things to China? Or skepticism from modellers over 3D-printing quality (better than injection moulded for newer SLA printing).

From some quick maths it seems like local production would be on-par for costs with cheaper material but higher labour costs and no shipping costs but while cutting lead times by i'd say ~75-80%. Unless my costings are way off because I've only printed some scratch-built models for myself and nothing commercial.

Shane.

[Noel]

3 hours ago, ShaneC said:

Just trying to understand the logic behind producing the models in China / Asia using injection moulding vs. doing it locally using SLA printing.

It seems sort-of crazy to spend €20-50K (if blog posts from British producers are to be believed) for runs of 500-1000 locomotives on moulds alone and having 12-18 month lead times on new products. Is it to do with labor costs associated with the painting & assembly that more pushes things to China? Or skepticism from modellers over 3D-printing quality (better than injection moulded for newer SLA printing).

From some quick maths it seems like local production would be on-par for costs with cheaper material but higher labour costs and no shipping costs but while cutting lead times by i'd say ~75-80%. Unless my costings are way off because I've only printed some scratch-built models for myself and nothing commercial.

Shane.

Hi Shane. No idea of the economic production cost figures, but I’ve never seen SLA remotely as crisp and sharp for fine surface detail as IMP. I’ve seen small items such as signals, small scenic structures, LLPs, etc, done in SLA that are passable once painted properly, but not yet seen a half decent coach or loco. Perhaps some of the Accurascale folk might be able to give you some feedback.  Noel

[Sails]

I suppose we have already seen this with Rails and their limited run on their printed vans last year. So its a viable option for sure, maybe the printing suits certain rolling stock. 

[Mayner]

JM Design is currently testing the concept of producing rtr rolling stock using 3D SLA printing with our first production batch of CIE Brake Vans due to arrive from the printers in approximately four weeks. 

We considered commissioning rtr plastic injection molded wagons from China but decided on SLA printing as it would have been uneconomic to produce the Brake Van as a plastic injection model.

While the majority of Chinese factories have a minimum order quantity of between 1500-2000 models, I would have had to sell far more than the minimum quantity to break even let alone make a profit.

The lead time from starting 3D design work on the project to manufacture has been similar to that involved in commissioning a plastic injection model, the design costs using a professional designer are similar the main difference is that the tooling cost is absorbed into the unit cost of the model rather that an upfront cost for traditional tooling.

I had planned to use a locally based 3D printing business for the production printing, but have had to outsource production printing to a company in China as it could not be done economically locally. I have a good working relationship with a locally based 3D printing business that carries out prototyping at a reasonable rate but was reluctant to quote for the production phase.

The 3D print is just the starting point, there is also the issue of painting and lettering the model, sourcing components such as wheels couplings and bolts from suppliers and manufacturers on three continents and the little matter of packaging the finished model.

To summaries

1. Commissioning a plastic injection model from China for less than 1500 models is likely to be un-economic unless you can charge a considerable premium.

2. The lead time to produce a decent SLA printed model is likely to be similar to commissioning a plastic injection model.

3. It may be necessary to outsource production printing, painting, assembly and detailing in order to produce a model at a competitive price.

 

Edited March 8, 2021 by Mayner

[Warbonnet]

6 hours ago, ShaneC said:

Just trying to understand the logic behind producing the models in China / Asia using injection moulding vs. doing it locally using SLA printing.

It seems sort-of crazy to spend €20-50K (if blog posts from British producers are to be believed) for runs of 500-1000 locomotives on moulds alone and having 12-18 month lead times on new products. Is it to do with labor costs associated with the painting & assembly that more pushes things to China? Or skepticism from modellers over 3D-printing quality (better than injection moulded for newer SLA printing).

From some quick maths it seems like local production would be on-par for costs with cheaper material but higher labour costs and no shipping costs but while cutting lead times by i'd say ~75-80%. Unless my costings are way off because I've only printed some scratch-built models for myself and nothing commercial.

Shane.

Hi Shane,

1. Labour costs

2. Expertise in producing highly detailed models 

3. Better fidelity 

If it was a crazy way to do it, nobody would do it. Nobody does runs of 500 for a mass produced China model mind you.

The 12-18 month lead time is not just in the China end of manufacture, it's also CAD design, artwork, survey etc. which is performed locally. Nothing China can do about that. SLA printing will of course get better but it will never replicated the highly detailed models we get from China on a mass produced scale simply due to the large amount of assembly (labour) that goes into making these models. It would still be assembled in China, or a similar country labour cost wise.

As an exercise, put the Murphys 121 and a OO works locomotive next to each other. Which has more detail parts? Which has more features? Which model is the better runner? Which is the better model? Which one costs more? 

If China wasnt the answer nobody would be using China for manufacture. 

Hope this helps!

Fran

[David Holman]

 A friend of mine at the Chatham Club does his own kits and is steadily broadening his portfolio. Recently asked him about the cost of producing a kit and have copied his comments below:

" If I was to do a run of 20 kits to sell to whoever and was sure I could sell 20, the price per kit would be around £275 - £300.  If it was more like a dozen would sell, the price would have to go up to nearer £400 and if only five, more like £700. This would be a complete kit with all but couplings included.  The etchers and casters are all running way behind at the moment and quoting times like 3 to 6 months for a new etch to be processed and samples produced.

  It helps if there was another loco that used similar parts, or shared the tender, this is often a good way to make something viable as 2 kits can be produced with only 1 1/2 times the work. "

 

 So, the moral of the story seems to be the more you sell, the cheaper it becomes, while to risk to small volume producers can be considerable unless they are sure they can sell what they produce.

 

 

•  
• Quote

[DJ Dangerous]

Labour costs have to be the main reason.

More then ten times cheaper in China than the US, according to this piece.

Imagine how much each model would cost the consumer if the labour costs in producing it were ten times higher, or more...

[BosKonay]

7 minutes ago, DJ Dangerous said:

Labour costs have to be the main reason.

More then ten times cheaper in China than the US, according to this piece.

Imagine how much each model would cost the consumer if the labour costs in producing it were ten times higher, or more...

Exactly that DJ, a well paid, highly experienced assembly technician (10 years experience of assembly model railway items) might earn $150-200 a month. During that month, they can assembly probably thousands of models / parts.

Every time I hear cries of 'why can't we make it here' compare that (highly trained and experienced) persons cost to the Irish or UK minimum wage, then multiply it by the 600 or so assembly technicians IRM / Accurascale use alone....

[DJ Dangerous]

2 minutes ago, BosKonay said:

Exactly that DJ, a well paid, highly experienced assembly technician (10 years experience of assembly model railway items) might earn $150-200 a month. During that month, they can assembly probably thousands of models / parts.

Every time I hear cries of 'why can't we make it here' compare that (highly trained and experienced) persons cost to the Irish or UK minimum wage, then multiply it by the 600 or so assembly technicians IRM / Accurascale use alone....

 

Ah, I'm sure people would have no problem paying €1900 per A Class instead of €190 if they were made in Ireland.

Let's make Ireland great again and bring manufacturing home!

[Bumble_Bee]

3D printing isn't as cheap as its touted to be honestly especially when you want high fidelity stuff. Even then, its never ever going to be as smooth as injection moulded plastic without a lot of work. Yes there are ways to do it (acetone baths etc) to smooth print lines but it also smooths edges and corners and details. 

 

3D printing is fantastic for large scale things like costume props (I regularly use 3d printed kits to make star wars blasters etc), but for something small like 00 gauge models its just never going to be clear enough. 

 

Whenever 3d metal sintering gets better though, THAT will be better than injection moulding because that could have similar levels of detail to die casting

[ShaneC]

29 minutes ago, Bumble_Bee said:

Even then, its never ever going to be as smooth as injection moulded plastic without a lot of work. Yes there are ways to do it (acetone baths etc) to smooth print lines but it also smooths edges and corners and details. 

Have a feeling you're talking about FDM printing here not SLA. Pretty much all the latest SLA printers print at resolutions as low as 50-55 microns (20th of 1mm).

You'd be hard pressed to spot a layer line with the naked eye on a Z-scale model (see-below) at that resolution let alone a OO-scale model - no sanding or acetone baths required:

(image from reddit)

11 hours ago, Warbonnet said:

Hi Shane,

1. Labour costs

2. Expertise in producing highly detailed models 

3. Better fidelity 

If it was a crazy way to do it, nobody would do it. Nobody does runs of 500 for a mass produced China model mind you.

The 12-18 month lead time is not just in the China end of manufacture, it's also CAD design, artwork, survey etc. which is performed locally. Nothing China can do about that. SLA printing will of course get better but it will never replicated the highly detailed models we get from China on a mass produced scale simply due to the large amount of assembly (labour) that goes into making these models. It would still be assembled in China, or a similar country labour cost wise.

As an exercise, put the Murphys 121 and a OO works locomotive next to each other. Which has more detail parts? Which has more features? Which model is the better runner? Which is the better model? Which one costs more? 

If China wasnt the answer nobody would be using China for manufacture. 

Hope this helps!

Fran

Interesting info, thanks for the reply! Are you able to say how much of your costs for each model goes on the cost of the moulds/material vs. labour/assembly?

I guess it's the painting time that really adds the costs.

Edited March 8, 2021 by ShaneC

[Bumble_Bee]

2 minutes ago, ShaneC said:

Have a feeling you're talking about FDM printing here not SLA. Pretty much all the latest SLA printers print at resolutions as low as 50-55 microns (20th of 1mm).

You'd be hard pressed to spot a layer line with the naked eye on a Z-scale model (see-below) at that resolution let alone a OO-scale model - no sanding or acetone baths required:

(image from reddit)

oh wow! thats impressive

 

Fair enough, I eat my hat, I wasn't aware that there were different types of plastic printing available yet

[BosKonay]

It depends a lot on the model and size of run but tooling is roughly 20-35% of the total. 
 

3d printing can’t come close to the cost at scale. 

[jhb171achill]

I noticed a while back that Shapeways offers several Irish models. Has anyone any experience of them, in particular coaches?

Gut feeling suggests to me that they're pretty crude, but maybe someone here has first hand knowledge?

[murphaph]

Didn't Noel get his printed 121 from Shapeways? @Noel

[DJ Dangerous]

21 minutes ago, jhb171achill said:

I noticed a while back that Shapeways offers several Irish models. Has anyone any experience of them, in particular coaches?

Gut feeling suggests to me that they're pretty crude, but maybe someone here has first hand knowledge?

 

9 minutes ago, murphaph said:

Didn't Noel get his printed 121 from Shapeways? @Noel

Noel did a great job with that one, but it still ended up looking like corrugated cardboard, despite his Trojan work.

[Noel]

13 minutes ago, murphaph said:

Didn't Noel get his printed 121 from Shapeways? @Noel

Yip, while I was delighted to build/kit-bash a number of them, they still were nowhere near IMP for surface detail quality. All sound equipped and love running these locos.

CIE 122

B121

In build

 

[Galteemore]

Duck test, 2’ rule. Passes both for me! Nice work Noel. 

[jhb171achill]

58 minutes ago, Galteemore said:

Duck test, 2’ rule. Passes both for me! Nice work Noel. 

Noel, they've served you well, and they look great weathered.

[brianmcs]

got this 47' flat from shapeways recently. it will easily pass the 2ft test. , though it does not have large flat panels like a coach would have.

[Noel]

9 minutes ago, brianmcs said:

got this 47' flat from shapeways recently. it will easily pass the 2ft test. , though it does not have large flat panels like a coach would have.

As an experiment I got one of these last year also shapeways but it was FUD (Frosted ultra detail), it runs very well on bachmann bogies. From memory it was designed by one of the RMweb guys Neil. I was pleasantly surprised how detailed and smooth it was for 3D. Love the big round buffers.

It runs freely and looks the part. I didn't order any more as I had ordered IRM versions at the time and 42fts are a bit modern for my preferred era anyway (ie pre containerisation or just at the start of transition). FUD doesn't seem to suffer warping.

[DJ Dangerous]

28 minutes ago, Noel said:

As an experiment I got one of these last year also shapeways but it was FUD (Frosted ultra detail), it runs very well on bachmann bogies. From memory it was designed by one of the RMweb guys Neil. I was pleasantly surprised how detailed and smooth it was for 3D. Love the big round buffers.

It runs freely and looks the part. I didn't order any more as I had ordered IRM versions at the time and 42fts are a bit modern for my preferred era anyway (ie pre containerisation or just at the start of transition). FUD doesn't seem to suffer warping.

 

Any close-up side by side photos with an IRM 42' Flat, for comparing the level of detail?

[BosKonay]

[Noel]

31 minutes ago, DJ Dangerous said:

 

Any close-up side by side photos with an IRM 42' Flat, for comparing the level of detail?

There is no comparison really as you can see from Stephen's post. The IRM ones have been fully decorated and include running numbers, etc, also more weight.  It was a bit of a PITA to have to source separate bogies and couplings as well as paint and decorate the FUD print, but it ran well and passed the 'duck test'. As I said I bought a rake of IRM ones anyway.

[murphaph]

The huge advantage with the IRM type flat wagons of course is that they are not plastic at all but die-cast metal, so they are the correct weight even unladen. Hiding a weight inside a plastic skeletal wagon, regardless of manufacturing method, is nigh on impossible so you end up needing to place a load on them, missing out on all that detail visible through the chassis from above.

[DJ Dangerous]

38 minutes ago, BosKonay said:

 

So, when are the 47' Flats hitting the shelves?

[Mayner]

The main draw back with Shapeways seems to be that models like the MGWR 6 wheel coaches only appear to be available in White Natural Versatile Plastic which has a surface texture like rough cast plaster  https://www.shapeways.com/product/QYWUZUKGH/0-76-mgwr-6w-brake-3rd-coach?optionId=96599245&li=marketplace

Some models like the 47'6" flats are available in Frosted Ultra Detail (FUD) but the material is more expensive, has a layered effect and quite brittle, I managed to shatter some FUD parts.

A superior surface finish without a layered effect should be achievable with Shapeways SLA materials, but a lot depends on the skill and experience of technician who sets up the CAD file on the printer and how the print house ensures that the model is not damaged during the clean up process and removal of the temporary support structure. https://www.shapeways.com/materials/sla-accura-xtreme/.

The goods brake is a good example of the standard of surface finish and and level of detail that can be achieved by SLA printing, the wavy roof and coupler sag has been resolved in the final design.

[ShaneC]

I believe Shapeways use SLS printing for their 3D prints which is a method of printing that sinters layers of plastic powder together using a laser. Has the benefit of being fast for reasonable levels of quality but it results in that grainy type texture you see on the models from them.

Edited March 9, 2021 by ShaneC

[BosKonay]

We’ve used SLA printing for test samples in the past but the cost is about £3000 for a non working locomotive. 

[ShaneC]

1 hour ago, BosKonay said:

We’ve used SLA printing for test samples in the past but the cost is about £3000 for a non working locomotive. 

That's painted and assembled as well I'd imagine, not just the body printed?

Labour shouldn't be that much more in Ireland vs. China.

I don't know where IRM has their production but more often than not the big US & European mode producers have factories in Dongguan - a city with very high wages for semi-skilled/skilled workers, a long on-going worker shortage and high safety standards compared to other Chinese regions. Wages there are only 40-60% below Irish wages for semi-skilled workers, granted they probably work more than 40 hour weeks. But it's not like wages are a tenth of here.

Edited March 10, 2021 by ShaneC

[DJ Dangerous]

2 hours ago, ShaneC said:

That's painted and assembled as well I'd imagine, not just the body printed?

Labour shouldn't be that much more in Ireland vs. China.

I don't know where IRM has their production but more often than not the big US & European mode producers have factories in Dongguan - a city with very high wages for semi-skilled/skilled workers, a long on-going worker shortage and high safety standards compared to other Chinese regions. Wages there are only 40-60% below Irish wages for semi-skilled workers, granted they probably work more than 40 hour weeks. But it's not like wages are a tenth of here.

 

Average wage in Dongguan is about a quarter of a Dublin wage, according to this.

Probably closer to a fifth or a sixth when factoring in the weekly hours, as you pointed out.

Still a huge huge saving if a project is likely to involve hundreds or thousands of man-hours labour.

Plus the plant and the experienced technicians are already in place, avoiding start-up costs and quality issues due to inexperience.

It would take a very bold investor with a lot of patience to start manufacturing in Ireland, and a market willing to pay ten times the price per item.

Maybe with time, as Ireland becomes reknowned as a manufacturing capital, more companies contract Irish manufacturers, and economies of scale kick in, prices would drop.

Possible, definitely possible, but probably not probable.

 

 

[ShaneC]

1 hour ago, DJ Dangerous said:

and a market willing to pay ten times the price per item.

I just don't think this would be the case - you've got for example Micro-trains in the US who manufacture in the US and their model RRP are less than most Irish / British providers (even accounting for VAT). Simply put - any profit lost per unit is made back in being able to bring new models to market faster / having more releases per year with the same amount of capex.

Regardless I understand the reasons why makers outsource oversees and thanks to all the answers given here! Some really interesting info.

[Mayner]

A direct comparison of a SLA print and a plastic injection molded model a test print of a GSR grain wagon and a Ratio LNWR coal wagon.

We went a bit heavy with the rivet detail on this wagon, but otherwise I am quite happy with the model and CAD work.

The blemish on the side of the grain wagon is I tried to apply dry print lettering that is past its sale by date, we tried to push the boundaries on detailing with this model but ran into problems printing the hopper release gear which will have to be beefed up. The challenge at the production stage will be to avoid removing detail such as brake levers and linkages while removing the support structure at the factory.

 

Although SLA printing appears to be a viable option for the small scale manufacture it remains to be seen whether its a viable option for the manufacture and assembly of rtr models.

The next two months shall certainly test the concept for production of the brake van!

The Chinese OEM factories that produce small production runs of <10,000 unit orders are only viable because they have the capability to take on and process orders from a number of clients simultaneously, a number of the companies that specialise in smaller production runs were set up by former employees when Kader decided to shut down a lot of its OEM manufacture to concentrate on its own in house brands about 10 years ago.  This left Hornby and other competitors stranded until the present generation of Chinese OEM model railway manufacturers entered the market.

I doubt that similar model railway manufacturing capability or experience exists in the West today

 

Edited March 10, 2021 by Mayner

[Warbonnet]

5 hours ago, ShaneC said:

I just don't think this would be the case - you've got for example Micro-trains in the US who manufacture in the US and their model RRP are less than most Irish / British providers (even accounting for VAT). Simply put - any profit lost per unit is made back in being able to bring new models to market faster / having more releases per year with the same amount of capex.

Regardless I understand the reasons why makers outsource oversees and thanks to all the answers given here! Some really interesting info.

We know and see the numbers involved daily. It is a fraction of the price to make these models in China when there are so many man hours involved making them. If the price was competitive we would be making them locally. However, like ALL our competitors we know the only place to make highly detailed model trains for a reasonable price is China. 

7 hours ago, DJ Dangerous said:

 

Average wage in Dongguan is about a quarter of a Dublin wage, according to this.

Probably closer to a fifth or a sixth when factoring in the weekly hours, as you pointed out.

Still a huge huge saving if a project is likely to involve hundreds or thousands of man-hours labour.

Plus the plant and the experienced technicians are already in place, avoiding start-up costs and quality issues due to inexperience.

It would take a very bold investor with a lot of patience to start manufacturing in Ireland, and a market willing to pay ten times the price per item.

Maybe with time, as Ireland becomes reknowned as a manufacturing capital, more companies contract Irish manufacturers, and economies of scale kick in, prices would drop.

Possible, definitely possible, but probably not probable.

 

 

This is nail on the head, right here.

Also, Ireland and the UK has never had the experience of making highly detailed model trains. Indigenously produced models in Britain were low on detail, somewhat crude and formed of few parts. Market demand in higher detail and better quality was another driving factory to China production. Another point is that many factories are looking to or have moved out of Guandong to other parts of China to seek lower costs including local taxes so they can be more competitive. 

The only way to produce model trains competitively here on such a scale is to bring minimum wage down to 20% of what it is now, and who will work for that? Also, you would have to keep your factory working 52 weeks of the year. That is a lot of models to produce to keep your costs competitive. We would struggle to achieve that, even with Accurascale factored in.  

[Broithe]

I once made a vegetable plot for someone who wasn't the most agile person. I suggested that we could make a watering arrangement for it that would only require her to turn a tap on and off now and then. The idea was to use one of the oscillating sprinklers that will cover a rectangular area, thus we wouldn't waste water. The area can be adjusted by blocking nozzles for two sides and by adjusting the drive linkage from the turbine-driven gearbox for the other two sides. Subject to it not being too windy and the water pressure being fairly reliable, then the area watered could be quite accurately controlled.

In those far-off days they were generally about £30, if you were lucky, and not widely available. I suggested that we keep an eye out for 'the right one' and we didn't need it straight away. The next week, she announced that she had bought one that she'd seen in a garden centre - "How much?", I asked, hoping that she had got the right sort, as she hadn't seemed to understand what I had in mind - "£2.99" was the answer. I admonished her for buying the wrong sort, imagining that she had got a cheapo rotary sprinkler, but, no, she had purchased a "Chinese copy" that was 99% as good as the European item that i had in mind.

I doubt that we could have sent the empty box back to China for what she paid for the whole thing, even if we had flattened it.

Each nozzle was an individual brass fitting, pressed into the main tube - and there was an attached tool supplied to deal with any blockages.

[skinner75]

It also helps cost-wise if you don't have to spend the money on R&D of a product