First 3D metal Printed Part

[JasonBModerator @jasonb]

I’ve been waiting a while for a suitable part to come up that I could get 3D metal printed just to see how it would turn out. The trunk guide for my next engine seemed to fit the bill so an STL file was created from my Alibre 3D CAD model and uploaded to a few online printing sites to see what the costings and options were.

I have seen mention of Craftcloud on a couple of other ME forums and ended up going with them. Stainless Steel 316L along with Aluminium seem to be the most common and cost effective materials and Craftcloud offer several prices for both, they also offer DMLS which is said to have smaller layer height than SLM which most other sites do so I went with a 316L DMLS part, basic sandblasted finish.

Production was quoted as 3-4 days and a further 3 for delivery and that is what it took. Several update emails to tell me of progress and good tracking all the way via Fedex.

I’m quite happy with how it turned out, better than most sand castings and as good as many lost wax. There was a tiny depression on the flange but as I had allowed 1mm machining allowance on that, the feet and the bore it was not an issue. I can’t see any layer lines and there is just a hint of a bit of remaining support in a couple of places. For an idea of scale it is 61mm long.

Milling (carbide) and boring (HSS) were fine, just a bit more pressure needed when drilling the holes (HSS) so no problems with machining which some of the earlier sintered 3D prints gave problems with.

Cost wise it was a bit expensive compared with what I could have fabricated it for or even cut from solid on the CNC. It was a fairly even split between the minimum job cost and postage costing just over £50. However having later looked at the Fedex shipping that postage cost would have covered anything upto 1kg and I could have got more prints done and still been within the minimum so could actually have gotten seven printed for the same cost bringing the individual cost to about £7.50. If using again I would group a few parts together so that they were all covered by the one off job cost as that would be far more economical.

M-Machine have just delivered the other “raw castings” so can you guess what it is yet?

[Craig BrownParticipant @craigbrown60096]

Very impressive

[Roderick JenkinsParticipant @roderickjenkins93242]

Oil field engine?

Rod

[Jim NicParticipant @jimnic]

That looks like the future of modelling for sure.

As to what it is, the closest engine I have to your trunk guide is a Stuart 10H.  However the Stuart is a little smaller and I can’t imagine you would bother with something as simple as a 10!

Something along those lines though.

Jim

[jimmy bParticipant @jimmyb]

Impressive.

I’ve looked/considered this route a few times.

Certainly gives me confidence in trying. A quick PLA print to prove size etc would give confidence in the process.

 

Jimb

[JasonBModerator @jasonb]

I suppose Rod is the closest as PM Research do make an oil filed engine as well as their #6 is said to be based on this engine.

Mine will be a half scale of the 1/4HP A J Weed engine of 1895 that was sold in several sizes from 1/8 to 2 HP by Chas A Strelinger and Co. I’ve not seen the PMR drawings and have just modeled mine from some videos of the restoration of one of the engines, this is a screen shot that I took my sizes from combined with quite a few that are given in the old Strelinger catalogue

Watch out for a build thread coming to a forum near you soon!

[Julie AnnParticipant @julieann]

A very interesting thread; the part looks good and I am impressed by the way it machined.

I have been experimenting with 3D metal printing over the past few months. Items intended to replicate castings are surprisingly accurate, although finish and small hole sizes are worse than for an investment casting.

More recently I have ordered some basic shapes in order to test mechanical properties such as surface finish, density, ductility and machining properties. The results will be written up in an article for consideration by ME&W.

Thus far I have used PCBWay. Pricing seems to be a little odd, all parts, of varying size and height, were around $30 each. A downside is having to pay in US dollars, leading to a lousy exchange rate on a credit card. Other payment methods are offered, but all have charges in addition to the part/shipping cost. I have a US dollar account but the additional charges made it more expensive than using a credit card. Craftcloud looks interesting, although I can’t see where they are based. They seem to be an intermediary, ie, they connect buyers with manufacturers but don’t make parts themselves whereas I think PCBWay make the parts themselves?

Julie

[JasonBModerator @jasonb]

They seem to me to be a German based “agent” for a large number of printers, foundries and CNC houses all around the world. depending on the material and methods chosed they will offer a selection of prices. You can then pick which suits your need and pocket. one of the first things you have to do is confirm your country, then it prices in pounds, I used paypal to pay and the account was “All3DP GmbH”

Worth uploading a file or two and trying some options. I could have also had that part brinted in bronze, investment cast quite cheaply from India or had higher resolution metal prints done for prices upto £750

[DalboyParticipant @dalboy]

Printed part looks good. I have yet to try making parts by fabricating from stock material and silver soldering, so way beyond the realms of 3D printing

[BazyleParticipant @bazyle]

We have some people in the Gaugeone3Dcircle experimenting with metal printing mostly using Craftcloud. Sometimes they have problems with files and details but generally promising if it gets cheaper. Mostly the shipping is as much as the part and often regardless of size. It helps to cluster multiple small parts onto a sprue which is then treated as a single item but if things get too flimsy it can cause problems and they might reject the job.
Research how the metal printing works and design accordingly. The post print separation and cleaning is rather physical and can cause damage. Small parts breaking off will cause the supplier big problems so they charge extra or reject a job that has flimsy bits.
Also unused raw material cannot always be recycled so gets a charge. You might find a house brick is cheaper than a drawing pin because of these extra influences.

[JasonBModerator @jasonb]

In the build thread of this engine Adrian said

On 25 March 2025 at 07:22 Adrian R2 Said:
Looks more aerospace than Victorian. With what you now know about 3D metal printing, could this base have been economically produced in a batch with the slide, and possibly other parts?

So I have had a bit of a play.

If printed the bed would be better hollow as you would get on a casting so I used the “Shell” function in Alibre to remove the bottom face and leave all the walls 3mm thick, all holes were deleted. I then created a STEP file to upload and these are the costings, all include taxes and delivery

Aluminium SLM Printing

1off £75, 2 off £66ea, 5off £59ea and 10off £57ea

Aluminium DLMS Printed (same process as the trunk guide above)

1Off £135, 2off £107ea, 5off £92ea and 10 off £87ea

Aluminum CNC cut from 6062, Did not have the hollowed out underside for this as it would be an unnecessary cost

1off £128, 2off £95ea, 5off £87ea and 10off £84ea

By way of comparison, the Stuart No8 Sole Plate is about the same size and with VAT and delivery would set you back £90.50. So the SLM would be cheaper and likely to be as good a finish as your average ME casting. paying a bit more for DLMS or CNC is not massively more and if you grouped together with friends or sold on the extras all could be a similar price to a commercial casting.

Obtaining the weight from Alibre which came out at 420g for the hollowed out version I could probably get change out of £30 if I made a pattern and had it cast in iron and posted back to me.

 

As for grouping parts together, I tried that with a few linked by a simple sprue

In Aluminium DLMS £132 for one print or in Stainless DLMS £220 again for one print. Quantities would give similar cost savings to the single part. might be cheaper if the trunk guide were horizontal as it would half the number of passes the machine needs to make though I’m sure at the other end they would group several peoples jobs together to get the best productivity out of the machine.

So probably not something you would want to use all the time but if the part could not easily be made by other methods it would certainly be worth giving metal printing or even CNC services a go particularly if you could make use of the multiples. Things like castings that would have used complex cores which would not be easy to get at with a milling cuttre or fine fins like you get on an aero engine would be the type of job that this method would suit.

[Adrian R2Participant @adrianr2]

Thanks – now thinking that if the short run economics reach parity with castings then there might be an opportunity for some enterprising supplier to hold a print-on-demand library as well as actual physical stock.

 

[JasonBModerator @jasonb]

Yes it would have some advantages, no capital tied up in castings sat on a shelf, quick lead times unlike castings, No need for storage.

Other plus points would be the ability to offer the product in different sizes so you could have one file for those working with a small mini lathe and a larger option for those with a 280-300 swing.

Taking it a step further, just sell the files. leaves the builder the option as to how they want to make the individual parts. Though with a bit of effort anyone with CAD abilities could redraw from printed drawings but there are those who fall at the first hurdle of not being able to use CAD.

[Adrian R2Participant @adrianr2]

CAD – indeed, I have experimented with a few widgets 3D printed in plastic by mail order but the cost of failure/redesign when it doesn’t quite come out as expected is acceptable. I haven’t been confident enough in my design-for-manufacture skills to risk trying metal yet.

[Norman MillierParticipant @normanmillier89056]

I have had quite a few things printed via CraftCloud now. This is the gauge one class 01 shunter I have been designing for some time. Plastic parts – chassis, bodywork and cab – I print myself, but the wheels and jackshaft weights are printed in stainless steel (wheels have been skimmed in the lathe), while the buffer beams, buffer heads and steps are printed in aluminium.

On CraftCloud the pricing is somewhat odd, for smaller items there must be a minimum price which means that often getting four or five of the same item costs the same as one.  Spruing items together is useful but I understand the detachment and cleaning process is pretty brutal so I make the sprues fairly robust, and I keep the items in similar orientations. When I have tried to have a large number of different items combined into one they have been rejected.

[JasonBModerator @jasonb]

They look to have come out well and probably a lot crisper than cast wheels would have.

I like the way that you have dovetailed the frame/buffer beam joints.

[Norman MillierParticipant @normanmillier89056]

I think the wheels worked out somewhat cheaper than castings would have been, and pretty clean.  The design is a mass of dovetails, which is fine when joining metal to plastic, but I think the surface roughness of the metal prints would make it difficult to rely on dovetails.

Generally I have used metal for parts which seem vulnerable, hence the steps and buffers.

I have done some wagon underframes in aluminium – along with the brake gear, axleboxes, buffer stocks.  They are a lot more satisfying than plastic, and the brake gear is far stronger. Easy to tap and screw together as well.

[MacolmParticipant @macolm]

I went to a visit at a metal printing firm a few years ago, but I have no idea how the technology has progressed recently.

The process I saw used a platten that could be very accurately lowered. A uniform layer of metal powder was dispensed, then a similarly accurate scraper levelled the surface to a very thin additional layer. Then a laser fused the powder to the previous layer where solid metal was required. The platen then descended by one increment (small fraction of a millimetre), and the process repeated with the laser fusing pattern modified as required to form the part. And so on. The un-fused metal powder just stayed in position, to be removed at the end of the part manufacture.

Run time was typically many hours, and most of the time seemed to be taken in depositing and levelling each layer, with the lasing pretty quick. Such a machine is costly, so the price to make a part is likely dominated by the run time. That being so, the thinner the vertical overall dimension, the lower the price should be. If a number of parts can be made side by side in parallel, it should cost little more than for a single part.

[petro1headParticipant @petro1head]

Interesting, did not know you could 3d print in metal

[petro1head]

On petro1head Said:

Interesting, did not know you could 3d print in metal

There’s an obligatory sintering process in addition to the printing.

[JasonBModerator @jasonb]

There’s an obligatory sintering process in addition to the printing.

 

Depends on the process. Some will print a mixture of the desired metal with binders and then the part is separately sintered afterwards. May even have a stage between the two to washes away some of the mixture of binders.

Others like this one use a laser to sinter the powdered material (without binders) which is what produces the printed part directly.

My part used the DMLS process. DIRECT metal laser sintering so no separate process

[Adrian JohnstoneParticipant @adrianjohnstone89946]

Actually DMLS, in spite of the name, has no sintering phase. It is also known as Selective Laser Melting (SLM) which more correctly describes the process: unfortunately the ASTM decided to group it with sintering processes hence the name confusion (see https://en.wikipedia.org/wiki/Selective_laser_melting for more).

The distinction is extremely important. 3D printing processes based on sintering create porous stuctures: the microscopic spheres of powder are heated to just below full melting point and then stick together creating a strong matrix with quite a lot of space in it. That’s good if you want strength coupled to low density.

In contrast to sintering, the SLM process liquifies the top powder layer which then freezes into the earlier layers, and you end up with something that behaves much more like solid metal. You can see the result in Norman’s photographs above: he was able to turn down the tyres on his 3D printed metal wheels and get shiny solid metal. I’ve had a chance to handle these, and they certainly feel like turned castings;  strong, high density and machinable.

As you can imagine, it needs very tight control of the laser energy levels, and a supply of metal powder with closely controlled grain size to make SLM work – too much power and you distort earlier layers; too little and you don’t achieve good flow. It’s an amazing techology now available at reasonable prices to hobbyists.

Adrian

[Adrian Johnstone]

It’s almost miraculous what can be achieved when computers start to manipulate matter, not just data.

On Adrian Johnstone Said:

Actually DMLS, in spite of the name, has no sintering phase…you end up with something that behaves much more like solid metal. You can see the result in Norman’s photographs above: he was able to turn down the tyres on his 3D printed metal wheels and get shiny solid metal.

Am I to understand from this that a 3D-printed metal object created using a sintering process (properly speaking) is visibly porous (like a Cadbury’s Crunchie) if you remove the outer layer(s)?

From my uninitiated viewpoint, additive-manufacturing processes that don’t place strict limits on the subtractive manufacturing processes that can be performed on objects after their creation (and DMLS appears to be one) are far preferable to processes that cannot tolerate much post-printing/sintering machining without compromising the strength or performance of the objects created.

[JasonBModerator @jasonb]

Probably as case of choosing the right printing process for the particular part and application. With modern CAD & FEA it is far easier to design a part and know what sort of loads it can take and make allowances for what the type of print can tollerate. Add in the ability to print a surface that is flatter than a traditional “cast” surface, include threads, etc and there is less need to machine. That is assuming things will be bolted together rather than using bonding methods and modern adhesives.

The video towards the top of this page shows the separate sintering method quite well and the video below shows the powder/laser method similar to how my part was produced except they are using titanium. Also shows the post processing of removing supports etc.

[Adrian JohnstoneParticipant @adrianjohnstone89946]

The level of porosity depends on the process and the powder grade, but it would be microscopic and not obviously visible to the naked eye. The raw surface texture is a litle different to a casting but can in some cases be polished or machined to a shiny surface.

In my limited experience, the sintered parts are isotropic and do not have a skin as implied by your question – not like a Crunchie… However in some applications surface treatment or just plain painting are used.

A colleague made some test pieces in sintered nylon a few years ago that were intended to be part of a low-pressure pneumatic system – that didn’t work as the parts were not airtight.

It’s worth noting that sintering of metallic powders using moulds, heat and pressure is quite an old technology: see https://www.alliedsinterings.com/guide-to-sintered-metal-products for instance. An advantage of the porous structure is that it can be oil-impregnated and thus be self lubricating – I use sintered bronze bushes in my Gauge One wagons for instance.

Adrian

[Author]

Posts