3D Printing rolling stock

mmyers05 Feb 25, 2012

  1. mmyers05

    mmyers05 TrainBoard Member

    137
    0
    8
    Okay, let me preface this by admitting that I have myself never used a Perfactory machine before; That said, I have spoken with several individuals who own and use EnvisionTec machines (outside the model railroad industry of course) and each of them told me that there are major practical "design limitations" to the types of shapes that can be printed using this technology.

    Perfactory machines are unique in that the parts 'hang beneath' the build plate and are physically lifted out of the build material as they are constructed. I have been told that, since this material is rather viscous, thin planes hanging parallel to the 'surface' of the build tank can warp or break simply under the force of this lifting action. In addition, the designer needs to account for surfaces on which the support structure can attach and "hang" the piece (so to speak). I've been told that it isn't too uncommon for a part to break off of it's supports and sink to the bottom of the tank if not properly designed.

    Once again, don't get me wrong, these machines look awesome! That said, I'm fairly certain that EnvisionTec has practical design limitations even if their PR department likes to tout that there are no theoretical "design limitations" (just like FUD or anything else). :)

    Also, in Mark's defense: if I have learned anything while researching/working around 3D printers - it is that the "substrate cost" is usually only a very small fraction of the actual cost required to make a part. I can't speak for EnvisionTec specifically, but for Z-Corp (Shapeways "Sandstone") for example, the powder substrate makes up well less than half of the overall cost (or so I have been told).
     
  2. Chris333

    Chris333 TrainBoard Supporter

    2,541
    252
    49
    Wonder what it would cost to print with that lime green stuff?
     
  3. robert3985

    robert3985 TrainBoard Member

    841
    57
    14
    Jason,

    I'm not missing the practical aspects. I don't have a problem with a businessman making a profit in his business. After all, what is owning a business for if not for making a handy profit?

    The EnvisionTEC rep has been very clear as to what the profit potential is for this kind of RP process, and has been willing to pledge EnvisionTEC's support to funnel actual RP customers my way to make "down-time" as little as possible and recoup the monies spent on the machine quickly. This is good salesmanship IMO.

    Actually, when I saw what Mark was charging, it was not a negative point for me. It let me know that, as opposed to Shapeways, there is a healthy profit potential in the Perfactory process. I am sure that he charges a lot more for quick turnaround customers.

    What is especially appealing to me, since I do my own investment casting, are the materials which will produce a precision wax for direct casting, OR the material that is hard and heat-resistant enough to produce vulcanized rubber molds, used to produce the waxes for investing, burning out and filling with molten brass.

    I'll be sending in a file for a demo printing next week on some complex and detailed parts that would be impossible to manufacture using the Shapeways process. I'll post photos here when I get 'em back.

    Cheers!
    Bob Gilmore
     
  4. daniel_leavitt2000

    daniel_leavitt2000 TrainBoard Member

    1,356
    21
    32
  5. Robbman

    Robbman TrainBoard Member

    1,141
    0
    27

    You need to take into account time spent modifying given files... it's not just adding supports, etc, but actually going into the stl and making sure no weird triangles are present that can interfere with a proper build. You also have other costs associated with a Perfactory machine than just the resin... build plates, projector bulbs, etc.

    You really should contact Mark... getting help and advice from someone who's been dealing with EnvisionTEC/Perfactory for almost a decade in this specific field is going to help you out much more than a sales rep.
     
    Last edited by a moderator: Aug 30, 2012
  6. mmyers05

    mmyers05 TrainBoard Member

    137
    0
    8
    I sent you a pm - apologies again for the delay
     
  7. wcfn100

    wcfn100 TrainBoard Member

    1,049
    63
    30
    I need to correct this post of mine as I was way off.

    The quote I got from Mark was 6 units for $140 USD.

    My apologies to Mark for not double checking before posting.

    In light of this, I hope to redo my part with him and have something usable vs. the Shapeways parts.

    Jason
     
  8. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    Every once in a while I decide to share the benefit of my experience and knowledge with others. Today I have chosen you, Bob! :)

    As Jason has subsequently posted, the quote was actually $23.33 each.

    That's what you get for believing the sales guy. Don't believe car salesmen either. In the context of model railroad parts there are numerous limitations, particularly if you want to have the surfaces free of supports that need to be removed and sanded or filed.

    Your assumption is incorrect. You could always ask me...

    Perfactory machines are capable of high resolution but don't be fooled by the Z axis figures quoted by machine manufacturers. It's generally irrelevant if X and Y resolutions are lower.

    ABS is a thermoplastic. Perfactory machines require photo-initiated thermosets. Envisiontec's claims re their materials are...optimistic. I know of a legal dispute re the company's claims vs actuality.

    You are misinformed. The curing in a Perfactory process is determined by a binary mask - black or white. No gray scale. They gray scale refers to the way in which the light from the lamp is modulated by the optical switch array to compensate for non-uniform intensity at the curing plane. This is a calibration matter. Layers are there - try building a part with a plane a few degrees off axis. Skillful part orientation can make this hard to discern but it is time-consuming.

    Hope you included shipping from Germany to NZ in that!
    The cost of resin (mine is $11 per cubic inch landed from Germany) is minor in comparison to other costs, but on that subject to the volume of material in the part must be added the supports and all the liquid material that clings to the part which must be scrubbed off after building - add around 40% more resin. Consider, also, the yield. For parts such as the diesel hoods that I sell it takes a number of iterations to get the supports right so that they are sufficient during the build but meet as little of the exterior of the part as possible. For some parts it can take up to 10 or 12 tries to get it right. This is not a factor for inkjet-style machines as the support is continuous and easily removed (with the downside being the marred finish). It's time-consuming as well. Time spent calibrating the machine, setting up supports and builds, trimming cleaning and sanding off supports, packaging and mailing is time not spent with my professional engineering clients. Fortunately I don't come cheap in that respect and can subsidise my model business! I do it because I enjoy it (mostly).

    ...and an ultrasonic cleaner and a UPS etc etc. Assume a capital outlay of around $108,000. Assume depreciation of 30% per annum and a maintenance contract of $10,000 (or that's what it was back in 2006).
    In this part of the world an unsecured business loan costs around 10%. There's $53000 of expenses in the first year before you even turn it on.
    Did the sales guy tell you about lamps? They are rated for 500 hours use and cost me 1600 euros each ex Germany. I can eke out around 1000 hours out of them so assume around $2 per hour running cost in lamps. At current layer thickness of 25 microns build rate is around 4 mm per hour.
    Did the sales guy tell you about basements? These are the silicone-lined trays that the resin sits in. They need replacing from time to time as the silicone disintegrates and the parts are deformed: $200 each landed. I could go on and detail electricity costs, insurance for a machine of this value, accountants fees etc...but I won't. The cost of resin comes way down the list after depreciation/finance, man-hours, lamps and maintenance.

    Given the above, an estimation of "profit" based on the volume of resin incorporated in the parts supplied seems a bit optimistic, don't you think?
     
    Last edited by a moderator: Sep 8, 2012
  9. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    Hey Robb! Good to see that you are still alive - we should talk some time!
     
  10. mmyers05

    mmyers05 TrainBoard Member

    137
    0
    8
    Hey Mark! It's good to see that you stumbled across our thread here!

    Thanks for clarifying things - I had heard that getting the supports correct could be a real issue - but you clearly know better than any of us. :)

    Just out of curiosity, are you as "size limited" as I have been lead to believe? More specifically, could you realistically print large steam locomotive boilers/fireboxes (think Northerns or small mallets for example)?
     
  11. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    I don't do so much stumbling these days...

    You can build most shapes with a Perfactory provided you are prepared to remove the supports and provided you are OK with the surfaces after you have removed the supports from them. If you want to preserve detail and surface finish the challenge becomes how to build parts with an absolute minimum of supports. There are various strategies for that such as rotating parts so that horizontal planes become 45 degree planes which grow "organically" or building parts "tall". The downside of the latter approach is that they take longer on the machine but it is generally worth the effort.

    You've heard the rumors too! :eek: Now it's all over the Internet!

    Without seeing the file I will state that boilers would probably build (not print! - that's for those squirty-type machines!) well - the machine is certainly large enough to do them in Z, HO and S. We top out at around 200 mm long. To minimise the effects of the supports there would probably be a split at say, the firebox, so that when the 2 parts are joined the 2 ends where the supports were "disappears". A registration feature and lap joint would insure a good join and alignment.
     
  12. daniel_leavitt2000

    daniel_leavitt2000 TrainBoard Member

    1,356
    21
    32
    Mark, while we have your ear, is there any progress on your CF7?
     
  13. Robbman

    Robbman TrainBoard Member

    1,141
    0
    27

    While I'm sure our beer over here isn't going to entice you to get back over here... I'm thinking our trains will :p
     
  14. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    I was wondering where that ear was. :)The CF7 project has ground to a halt as have a number of other N scale and Z scale projects. Thanks to "quantitative easing" and the state of the US economy and their effects on the exchange rate the returns on model making just don't warrant my spending time designing new models for my range. To illustrate; when we moved to NZ one US dollar bought 1.77 New Zealand dollars. These days it buys 1.24. My call is that N scalers won't accept a corresponding price increase. I keep the present product range ticking over (at prices unchanged from 2004) because the design work and debugging is done. I still do a lot of design work but of a different type: later this weekend I shall be working on modifying the design of an automatic iron tapping/lancing/rodding/mudgun machine for the submerged arc iron furnaces at the local steel mill. It's interesting in its own way (it runs on rails and makes a lot of noise!). I am still doing some model design work but it is for one of the well-known US manufacturers who can cope with price increases. You will see the results in N and HO soon.
     
  15. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    Might be in Colorado next year - my sister is keen on a house swap with us. Nearest railfan location; Palmer Lake on the Joint Line. I have already mentioned it to Jason (he's fairly local to there). Of course we would really love to get back to Virginia (any high hoods still running on the Pokey??) but with a family of 5 we are looking at a serious budget. Mrs M4D will go on to the UK to see her elderly mother before she is too confused to remember her name (the mother-in-law, not Mrs M4D!).

    As for the beer; there's nothing wrong with a good American pale ale (or 2 or 3). Fond memories of live bluegrass and pale ale in venues around VA! There's a lot of Americana we miss.

    On the model front I have strayed; broke out the European HO collection a few months back and something cracked. Now I am busy buying stuff from the German Ebay and planning a layout.

    Are you still on the old gmail account?
     
    Last edited by a moderator: Sep 8, 2012
  16. mmyers05

    mmyers05 TrainBoard Member

    137
    0
    8
    Interesting... I my neck of the woods we call runs of the machine "builds" but individual parts "prints" (powdery-type machines to clarify) - I mean, they are still called "printers" after all. Then again, I do see the distinction that you are making... lol

    Thanks for the info regardless :)
     
  17. Mark 4 Design

    Mark 4 Design TrainBoard Member

    127
    0
    14
    Don't mind me Matthew; I'm a pedant. it's one of my hobbies.

    The Perfactory process is like a series of photographic exposures through a mask and thus has little similarity to a printer.
     
  18. cnw mike

    cnw mike TrainBoard Member

    128
    0
    8
    I have about two dozen ore cars primered, and another dozen painted boxcar red. I wanted to go ahead and finish one just to see how they will all turn out. I played around with a few different weathering methods. The grabs on the upper corners are wire, everything else is printed except trucks and couplers.

    [​IMG]
     
  19. cnw mike

    cnw mike TrainBoard Member

    128
    0
    8
  20. Robbman

    Robbman TrainBoard Member

    1,141
    0
    27

    Of course :p
     

Share This Page