Preping 3D printed parts

maxi-model

maxi-model

UK/US/ROW steam narrow gauge railways 1:1
27 Oct 2009
4,816
396
Bucks/Oxon/Northants area
I am in possession of my first kit that includes some 3D printed parts. My question is - how do you remove those "stratfication" lines caused by the nature of the layered print process ? They seem somewhat resistant to sanding and some of the parts I have would be impractical to sand due to their complex shapes with many narrow ressesses. I know there is chemical solvent solution but I do not know what it is or the risks of using it.

While I think i understand the benefits of 3D printing technology and its applications for making possible very limited run subject matter I feel it is a somewhat overused production method because of its novelty. Sure, use the technique to cut the cost of producing or simplify the creation of a master from which replicas can be made but if you can see a market to produce 30 + replicas, of a single item, then for gods sake provide your customer with the fully finished article. Create the replicas with a quick cure resin process instead. I rather resent the idea of having to do a lot of time consuming finishing work just because a vendor does not seem to wish to take the risk on producing multiple units up front or at least investing in the creation of a mold to produce them from. For the time being I shall be refraining from any buying any such article tha advertises it contains parts from that method of manufacture. OK, rant over. Max
 
M

Moonraker

Registered
25 Oct 2009
947
88
South Australia
Max,

I scratch build my own models using 3D printing and have to do the same clean up of lines as you. My solution is to go to my local auto parts store and buy a can of spray putty. This is used to fill car body scratches. Just spray it on and then rub it down with wet and dry paper or fine emery cloth.

However, rivets are always a problem. With a good 3D printer, they can be printed accurately to scale but then often get trashed in the cleanup process. My solution is to not print the rivets but to print holes instead where they are to be located. Then, last step before painting, buy a bag of brass rivets (Garden Rail Specialists used to sell them), drill out the holes with a Dremel and glue the rivets in place.

For flat surfaces I use laser cut acrylic which gives a perfectly smooth finish. Again, when drawing the shapes I include holes for the rivets. I use this a lot on diesel locos, wagons and coaches. On steam locos it is good only for the frames, cabs, tenders and buffer beams and other flat areas.

I am not sure that resin is a solution for locomotives. I have used it for buildings and could see it being used for steam loco backheads, domes and chimneys but not for hollow boilers and fireboxes.

Regards
Peter Lucas
MyLocoSound
 
  • Like
Reactions: mike
LGB-Sid

LGB-Sid

Registered
19 Sep 2016
1,129
540
61
UK
It also comes down to the Quality of the model the vender used and the settings he used to print it , speed verses quality and the orientation of parts. A lot of finishing work can be removed from prints but usually at a cost in print times which I suppose is why venders don't provide the quality that people are exspecting. It is a process that will always leave lines etc in the prints but with better planning and settings at the print stage and quality filament, they can be minimissed a lot. None of my models have needed any filler , just a few coats of primer and sanded with wet and dry. same as Max I dont print the rivits in either as they stop you getting the surface finish you want , I print holes as well :)
 
PhilP

PhilP

G Scale, 7/8th's, Electronics
5 Jun 2013
26,145
1,567
Tamworth, Staffs.
As has been said:
Quality starts at the design stage, and knowing how the material acts, and how the printer lays it down..
Choice of material, and print-speed, can make a big difference to quality, as can orientation of the part being printed.

Finishing:
It used to be recommended you 'burnished' the item with a hard round-ended stick. - To reduce striations and give a smoother surface.
There is (at least) one solvent solution that can be used. - Though it can soften the print, and (in my experience) takes ages to evaporate/dry/sort-of harden-off again)..
Spray 'putty' (also called spray filler) can be useful, but you may 'smother' detail, if not careful.

Suitability:
The jury is still out, on whether these materials / techniques will be long-lasting enough??

i fear many of the printed materials are unknown quantities, for life, and stability..
Leave a Stainz out in the sun all afternoon, and it will get hot. Leave a 3D print out in the sun, and it will probably turn into a work of art!
Some filaments are not plastics.. How long will they last?
How will the 3D print material age? - Will it degenerate, becoming brittle?

Resin prints:
Give a better finish, straight from the printer..
But, the items are brittle. - They will fail a 'drop-test'
 
korm kormsen

korm kormsen

Registered
24 Oct 2009
2,547
80
Last edited:
M

Moonraker

Registered
25 Oct 2009
947
88
South Australia
Acetone softens ABS so you have to use it very carefully otherwise you will ruin your model.

In my experience ABS is not greatly affected by hot sun; something we have lots of in South Australia. However I did once have a problem with the 3D printed roof of a railcar. If it was in the sun for an hour or two the roof distorted a bit and curled up a bit at the ends. When I put it in the shade it would restore its correct shape in a few minutes.

Regards
Peter Lucas
MyLocoSound
 
musket the dog

musket the dog

Professional engineer, amateur modeler
31 Oct 2009
607
177
Leicester
nlrr.webs.com
I have also used Halfords filler primer (with the yellow cap) to get a good finish on printed parts. It goes on thick enough to fill in the lines between the layers and dries quite chalky so it can be sanded quite easily.

I'm not nervous about the longevity of the plastics. It has come up in a couple of other topics on the subject but the two most common materials used are ABS and PLA with are very commonly used in the plastic body panels of cars. Fair enough I think some of our models will probably outlast most of the cheap cars built today, but they'll probably get abused more. Until I knock it off the shelf, again.
 
Greg Elmassian

Greg Elmassian

Registered
8 Mar 2014
4,038
579
San Diego
www.elmassian.com
PLA is not heat resistant, and uncoated can decompose easily, in fact one use is temporary structures in surgeries to grow tissue and then it decomposes into lactic acid. Please research if you disagree on this point.

ABS is good on the other hand for our applications. Harder to use in filament printers apparently.

Greg
 
FurkaSOCal

FurkaSOCal

Registered
26 Dec 2017
95
41
29
San Diego
The best solution I have found is to first sand the model with 120 grit sandpaper, then spray on a filler primer. After the primer is dry, sand the model with wet sandpaper using finer grit each time. It takes a few layers of primer and paint but you will quickly not be able to tell the part has been printed.
 
LGB-Sid

LGB-Sid

Registered
19 Sep 2016
1,129
540
61
UK
PETG is another common filiment that you could use , withstands heat far better the PLA but not quite as good at ABS , but easier to print with as it's less prone to warping while printing. There are lots of other filiments now on the Market that are suited to various uses.
 
maxi-model

maxi-model

UK/US/ROW steam narrow gauge railways 1:1
27 Oct 2009
4,816
396
Bucks/Oxon/Northants area
I think the 3d printers need another order of magnitude (10x) of improvement before we see "quality" that matches the injection molded bodies we are comparing them to.
My thinking too Greg. I'm very familiar with quick cure resin component manufacture, a related but lower cost at low volumes compared to injection moulding, and the economics around that and other low volume production techniques such as photo etching and white metal/brass casting. I can see the point of using 3D printing, and other "on demand" processes if you cannot see a market of sufficient size to use more "conventional" manufacturing processes or to use them as a prototyping tool to generate part finished "masters" to employ as a basis to work from to provide a fully finished component(s) from other further processes.

Problem is I am seeing some makers seemingly using these "on demand" processes more to remove as much risk, i.e. minimise the possibility of dead stock, rather than properly assessing the potential market demand, still quite limited, and applying the correct production processes. The result at the moment are a glut of interesting models that require too much work and risk of screwing up on the part of the buyer to achieve a decent result. Some makers seem to have got tunnel vision and are misapplying the technology and end up offering a product that is difficult to realise a satisfying end result. How many people are not buying because they are looking at what is offered and saying, "not for me, too much work and risk to get a reasonable finished article" ?

The other thing coming out in this thread is how one proceeds with handling, finishing and storing a 3D printed component is down to the type of plastic used in the process. A total minefield for the buyer who is laying out money and taking the risk if that is not explained by the supplier. Max
 
musket the dog

musket the dog

Professional engineer, amateur modeler
31 Oct 2009
607
177
Leicester
nlrr.webs.com
I wonder if part of the attractiveness to suppliers planning larger batches is that you don't need to maintain your tooling? Ok, so you still need to maintain your printers, but you don't need to replace or refurbish moulds or dies.

I've had some very good resin castings as parts of kits or detailing packs but I've also had some really poor parts where it is obvious that the tool is completely past it. As much time as you save with a good surface finish, it's very frustating as a customer to be faced with a part with a large tool mismatch to overcome or flash thicker than the part itself. I have returned a couple of parts that were completely unusable (improperly filled moulds and huge amounts of bubbles), at least considering I had paid money for a kit. It did make me wonder what the reject rate would be like for resin mouldings if you were only sending out (Almost) perfect parts? How would that number compare to failed prints from a decent, reliable printer?

As was alluded to, I think everything is made much more difficult by not knowing exactly what you are getting. Some PLAs are good for food packaging, some will dissolve in organic acids, some will sit quite happily as the end tank of your car radiator up to 140 degrees (Celcius). Unfortunately the only way I can think to know exactly what you are dealing with without the information up front is some testing and some trail and error. I would imagine that anyone providing these components as part of a kit would be confident enough with printing to be using ABS (or maybe even PA) anyway.
 
idlemarvel

idlemarvel

Neither idle nor a marvel
13 Jul 2015
2,217
324
Ascot
I keep thinking about investing in 3D printing and this thread is coming up with some interesting points and ideas beyond the usual "what printer should I get". It seems that you need to invest quite a bit of time in trail and error with materials and print finishing before you can get acceptable results, but they are certainly possible even with today's technology. I'm watching FurkaSOCal FurkaSOCal efforts with interest, for example.
 
  • Like
Reactions: FurkaSOCal and mike
L

LGeoB

Registered
12 Dec 2017
121
22
Perth, Western Australia
I must admit my main use for a 3D printer is making replacement bits for LGB trains - buffers, that little scroll bit under the roof of the two axle coaches and some box wagons. I slap on some unthinned Humbrol enamel paint that covers up the grooves in the printing process. All looks good from the usual few feet.

Geoff
 
maxi-model

maxi-model

UK/US/ROW steam narrow gauge railways 1:1
27 Oct 2009
4,816
396
Bucks/Oxon/Northants area
I wonder if part of the attractiveness to suppliers planning larger batches is that you don't need to maintain your tooling? Ok, so you still need to maintain your printers, but you don't need to replace or refurbish molds or dies.
When I used produce a range of built and kit form high detail 1:32 slot cars, in the 2000's, I used a company based in Birmingham UK - WCM | . Their then boss was an avid modeler and had set up a section of his company to provide molding/casting services to artisan model companies, such as my own and others we see plying their wares in this hobby. They still do this work to this day. They are the best in the business.

They use a vacuum molding process that obviates the issues with bubbles and part filled molds. They offered an all up unit cost per molding with the cost of the mold amortised within that cost. With guaranteed perfect reproductions every time. They also offered an alternative of paying for the mold up front and then a lower unit cost per molding. You took the risk if you pushed the units on a single mold on that deal. Their estimate then, in 2005, was that a typical silicon mold life was around 50 shots +/- , life dependent on factors such as the severity of undercuts and fine detail built into the piece being molded. They offered various types of quick cure resins dependent on the application the final article was being put to. For "dynamic" model, such as a working train or slot car they would use a resin that was more shock resistant in its cured form, other types for models that were intended for static display only. This consideration extended to cast metal parts they could produce - They used pewter rather than white metal where a part could be expected to be taking a knock in use.

Most of the poor work one sees is usually due to the use of open backed molds that have had the resin pored into, poor pattern design and trying to get a mold to do just a bit to much work to pay for itself. Then there are those individuals who "back mold" others work for a profit, a degraded 2nd generation molding. It comes down to the old price/performance equation - 3D print, no physical master just a data file (different skills to a costly pattern maker), no mold just the cost of use of printer devise and consumables. Quality of finished article and possible need for further "fettling" by end consumer dependent on printer "resolution" (cost implication ?) - Resin molding, cost of master production (you could 3D print that and finish it to its final form if you have the skills or use a higher resolution to create this master), cost of mold with 50 unit life and unit cost for molding and material. Quality dependent on master utilized and processes used. But, there should never be a need for extensive finishing of the final product by the end user with this process.

And here is my issue - There must be a cross over point in price/performance where the the benefits of a purely "on demand" process, like 3D printing, and a small volume molded process using quick cure resins, which would mitigate toward the latter, to provide a more desirable and usable end product. The "it looks ok from a few feet" does not cut it for me. When I buy a model of any sort it must pass muster at a few inches, especially if it is sold as something that will build, with the right generally accepted skills, to a faithful scale model. As you have mentioned Rickey, in relation to molded resin product, in essence what we have being passed off sometimes as the finished article is something that cannot be fashioned into a reasonable looking component without a lot of time, effort, skill and modicum of risk. Essentially what we would "return to sender" under any other circumstance.

It seems the makers have been blinded by the science and seduced by the lower risk to their capital outlay by not having to shell out for molds with a finite life that they may or may not get the full use of. In exchange we might get subjects produced that may never have been economical before these pure "on demand" production processes were available but we may get inferior/unsatisfactory product because of an unnecessary blind adherence to the wrong production process. I cannot be the only person that is now saying no to anything that includes less than perfectly resolved 3D printed components. It's a bit self fulfilling - you 3D print because you think your subject matter will not garner the sales to risk the outlay for a mold. You detract from the sales potential that would justify the use of molded components because of the first decision. Or have I got this all very wrong.

Sorry, I think I have got it off my chest now :D
 
musket the dog

musket the dog

Professional engineer, amateur modeler
31 Oct 2009
607
177
Leicester
nlrr.webs.com
Thanks for taking the time to get it off of your chest Maxi, a very informative and intesting read. As it turns out I've used the same company for some of our prototypes at work, I can't believe I didn't put 2 and 2 together and missed completely that vacuum casting could be pointed the direction of models :oops:. As you say very useful for things like under cuts and internal detail that you might not think possible at all with normal moulding processes.

I don't think you've missed the point at all. I see lots of printed models that are structurally excellent but are lack the same quality in the final finish either because the builder doesn't know how to finish the material or isn't worried about it. At the very least the supplier should be letting you know exactly what you are getting and how to treat it. Some are quite good at this, others less so. I don't think putting up a picture of a professionally finished model with no warning of the extra work you're going to need to get it to that stage, is quite fair on the customer.
 
P

Paul M

Registered
25 Oct 2016
5,046
964
57
Royston
This is a very interesting discussion, I've been looking at some 3D printed kits, but like the original poster, i was concerned about how to get a good finish.
It would be interesting to return to this thread in a couple if years to see if the arguments still apply
 
  • Like
Reactions: LGB-Sid
PhilP

PhilP

G Scale, 7/8th's, Electronics
5 Jun 2013
26,145
1,567
Tamworth, Staffs.
It would be interesting to return to this thread in a couple if years to see if the arguments still apply
Yep! - We might have got round to starting the kits by then! :rofl::rofl:
 
mike

mike

Master at annoying..
Staff member
GSC Moderator
24 Oct 2009
50,080
2,050
Rossendale
www.gscalecentral.net
Buy your own printer.. Then the quality is on you..?
Share the 3d..everyone wins.