MY SD40-2 IN 3D

Absolutely amazing!

I bet if you upload the .STL files to Shapeways, you can sell a few of these to offset the costs of having yours built!  Plus, the added bonus of having some pieces, like the frame(s), but able to printed in stainless and/or other steels.

Cool!

Thanks,

- Mario

Very nice, but I don't think the drawing is 1/1 (or 1:1) scale. If it were, then the drawing would be the full size of the actual locomotive. What the image shows is that the young lady and the engine are both drawn to the same  scale, whatever that may be on the original drawing.

B Smith posted:

Very nice, but I don't think the drawing is 1/1 (or 1:1) scale. If it were, then the drawing would be the full size of the actual locomotive. What the image shows is that the young lady and the engine are both drawn to the same  scale, whatever that may be on the original drawing.

Actually it probably is 1:1 in the Sketchup modeling space.  As in if you opened the drawing in Sketchup and pulled a dimension (say bolster center to center), the result would be the actual prototype dimension (43'-6", presuming the model is drawn accurately).  A good rule of thumb in my experience is that the input (2d plans and elevations, or 3d cad models) are drawn at full scale, while the output (hardcopy paper drawings or O scale shells) is scaled to the desired size.  It's not an absolute, and things can always be converted back and forth with a bit of math, but I think it's a best practice in general.

The second conclusion is of course absolutely correct.

Jim

Great, definitely will be following this. I agree with Mario there will be people who will by your engine. The P 48 guys come to mind.

Jpv,

Don Strack's UtahRails website has a lot of useful information about the UP sd40-2's, including things like nose length notes, field measurements of example units, and EMD spec sheets indicating major dimensions and clearances.  The latter includes some section information including the frame and air reservoirs.  Hope that helps.

Jim

One other thing to consider is purchasing many of the required parts from Atlas as opposed to printing them.  I tend to find printing large parts to be expensive.  Purchasing injection molded parts may cost less per part, although postage to France may or may not ultimately negate the savings.  

Existing Atlas parts that could be used for this model include: sd40 shell, sd40 or gp60 (for Q fan era units) radiator and dynamic brake fans, gp60 or gp15 dash 2 type cab (although they have thick walls), 81" or 88" noses including high short hood versions (also with different headlight positions), cab sub bases, and non-dynamic brake hatch if applicable.  You could even stretch an sd40 frame if you wanted to take that on.  You could also use the Atlas fuel tank and air reservoirs, although I not sure the tank profile is totally correct and the air reservoirs are molded in.  Some cutting and rearranging of hood parts would be necessary.  And this all depends on parts being in stock at Atlas.  

sd40 chassis parts diagram

sd40 body parts diagram

gp60 body parts diagram

Shoot me a note offline if you want to discuss in detail.  This is a project I have been considering for a long time using Atlas parts.  

Jim

Thanks all for your comments!

I will try to answer each of you!

----> Mario: my goal is not to sell parts of this engine; it's only a challenge for me to make it. I will not give the .stl files to Shapeways because they are too much expensive and above all because I work with a guy near who makes for me all the pieces I want. If somebody wants to build this model or to have some parts of it, he can simply contact me and I will give him what he needs. However, I understand your thought and I thank you.

----> B SMITH: the 3D allows you to draw full size or scale size; the drawing I posted is 1:1 scale! you can see some dimensions I measured randomly:

---->SUZUKOVITCH: thanks to follow my madness! but I'm afraid when you say that people will buy!!!!!

---->Jim: thanks for Don Strack's website; I didn't know. Parts from ATLAS is another solution but the shipping costs for my country are more and more exorbitant!!!! main reason to build it in full; but it's sure that I can make mistakes on some parts; I only have photos and drawings; the best would be to have a real engine close to me  but it's impossible!  I try to do my best with what I find and it's not always easy; TRAINIAX's website is also a good information source and helped me often!

jpv in France

Thanks for the clarification about the possibility of using the 3D to draw at 1:1 (or any other scale) and to print or display parts of the file at 1:1. That makes sense to me, but I still don't quite understand why you refer to the small posted drawing as 1:1 scale. In my understanding of the terms, a drawing of the locomotive done at a scale of 1:1 would have to be almost 65 feet long, since 1 actual foot of the locomotive requires 1 actual foot on a drawing down at that scale. The posted drawing and associated computer file apparently allow you to display actual 1:1 dimensions between selected points on the locomotive, without the drawing itself being at a scale of 1 foot = 1 foot. I must be missing something in the discussion!

B, I've been trying to think of an elegant way to describe the scale issue, but I'm having a hard time coming up with something.  So I'll try to come about it from a couple of different angles and hopefully something will click.  But first I'd say that you are correct in that a 1:1 scale drawing of an sd40-2 would in fact be something on the order of 68 feet long.  The hitch comes with the second assertion, about the disconnect between the displayed dimensions.

Two ideas to keep in mind:  1) A person is often used as a reference for size to give a sense of how big things are, because we understand about how big or small an average person is.  So we can compare un-dimensioned things to people and get a sense of how big they are.  In these types of comparisons, the person is always actual 1:1 size in the vicinity of 5-6 feet tall. Like this:

That green one is pretty big!

2) Typically when we draw in 3d, or even 2d, in a cad program, we are drawing in an infinite model space that is akin to the 3d world we live in.  Things are typically drawn in "1:1" size, just like if I went out to a parking lot in the real world and drew 2d plans and elevations of an sd40-2 in chalk on the pavement, that were 68 feet long.  And next to it I built a 3d representation of the sd40-2 made from legos, that was also the full size of a real sd40-2.

The "drawing" posted above is the equivalent of this 3d lego model, meaning in the Sketchup 3d model space it is actual size, and if I measured it in that model space, it would be 68 feet long.

So considering these ideas, the thing to do is think of the image above as a representation of an sd40-2 rather than the object itself.  It is exactly the equivalent of taking a photo of a real sd40-2.  Whether the photo is a 3x5, 4x6, or 8x10, the loco in the photo is still a full size 1:1 object, and this is really what Jpv is saying above.

Regarding actual 1:1 dimensions being represented on a less than 1:1 drawing/photo/representation, think of it in the following way.  Let's say you take a broadside photo of an real prototype sd40-2, download it to your computer, and print out a copy of the photo on 8 1/2 x 11 paper.  Clearly the photo is not 1:1 scale, but rather something like 11" = 68' scale.  Now you go back with a tape measure to the real locomotive you photographed originally, and take a measurement of something, say 43'-6" between truck bolster centers.  Then on the photo you printed out you mark 43'-6" from bolster center to bolster center.  Clearly in your photo, the bolsters aren't actually a real 43'-6" apart, but rather about 7" apart,  This is exactly what is happening with Jpv's image above, except instead of measuring a real life sd40-2 he is measuring a 1:1 3d model of a sd40-2 in the Sketchup model space.

That's the best I can do for now.  I hope this answers your questions adequately, especially because I'm going to have to dig a lot deeper to come up with something better .  Let me know if you still have questions.

Jim

No, I must respectfully disagree. The "scale" of a drawing (or of a model) is determined solely by the ratio that the drawing (or model) bears to the actual artifact. A 1:48 drawing of an actual 40-foot boxcar will necessarily be 10 inches in length, because at a scale of 1:48, one inch on the drawing is understood to represent 48 inches (4 feet) of the actual object: Therefore 10 inches in the drawing represents 40 feet of the "real" boxcar. If the drawing is reduced in size (or expanded) on a copying machine, then it will no longer bear the same ration to the original and will  no longer be a 1:48 representation of the original object, and measurements taken from the drawing will have to be adjusted accordingly. If the drawing is enlarged 2X, then the ratio will become 1:24 rather than 1:48.

That's what a scale drawing is all about: it represents the original object at a uniformly and consistently reduced size. Think of architectural plans for a house, which are often drawn at a scale of 1/4" = 1 (1:48) or 1/8" = 1 foot (1:96). The usefulness of such a "scale" drawing is that it communicates necessary information to the contractor without the need for drawings that are as large as the finished house itself. The builder understands that at a scale of 1/4" = 1 foot, a 3/4" inch dimension on the drawing for a certain doorway means that the actual doorway shall be 3 feet wide. The "scale" of the drawing expresses a direct mathematical relationship between the actual physical object and its representation on paper, or on a computer screen. Determining the final or "actual" measurement of the doorway therefore requires a computation or translation, because it is not directly given by the drawing.

Now, in this case it seems to me that what we have is not a 1:1 scale drawing of the locomotive (which would have to be 84'-8" in actual length), but instead a graphic interface presented at some arbitrary scale (the exact scale is not important) which then allows access to the 1:1 data stored in the underlying computer file. In other words, the drawing itself is clearly not at a scale of 1:1, but it does allow direct access to the computer's file of underlying actual dimensions without the need for any further computation or translation. 

A thought experiment: Your architect draw plans for your new house at a scale of 1 inch = 4 feet, but then you go to Kinkos and make enlarged copies to give to your contractor (perhaps because you think they are easier to read). Let's suppose the copies you give to your contractor have been enlarged by 25%, but that the "scale" designated on the drawings still says 1" = 4'. The result will be that every window and doorway and so on in the actual house will be larger than intended, because every inch on the drawing that was intended to represent four actual feet has been expanded to 1-1/4", which represents 5 feet at the original, intended scale of 1" = 4 feet (1:48). 

The point is that you can't provide correct scale dimensions at 1:1 on a drawing that is not itself 1:1, any more than you can accurately represent dimensions at a scale of 1:48 on a drawing that is not itself 1:48. That's why hobby magazines print (or used to print) separate drawings scaled for O, S, HO, and N. If you don't know the scale of the drawing, then it's useless for building an accurate model in a particular scale, because you don't know how to "read" the drawing.

A cumbersome alternative is to print a drawing of the original object that gives only the actual dimensions in feet and inches, which does not have to be printed at any particular scale, then let the model builder himself make all the necessary mathematical calculations to reduce these actual measurements to their scale equivalents in whatever scale he chooses (O, S, HO, etc).

Your final comment about measuring a reduced 11-inch photograph of the original locomotive in which "the bolsters aren't actually a real 43'-6" apart, but rather about 7" apart" only serves demonstrate my point that the photo (or a drawing made from it) is not a 1:1 scale representation of the original object. By definition, a 1:1 scale drawing or other representation of the original artifact must be exactly the same size as the original, otherwise it is not 1:1.

Anyway for me the scale, the size, etc....the most important thing is that the computer gives me the dimensions of the real engine like it had been drawn by "Fat Bear".

The engine is displayed on my screen and the dimensions are those of the real engine; after I have scaled the model, the engine is displayed once again on my screen and the dimensions are once more the real dimensions (in 1:48 scale). That's all I want! no matter for me the size of the "image" on the screen.

Well, let me show you other parts of the engine:

long hood sup part 1:

long hood part 2:

long hood part 3:

front short hood:

central air intake and exhaust stack:

And the 2 sides that fit under the part in the first photo:

left side:

and right side:

Eyebolts, rivets, grab irons, etc...will be added once the 3D parts done.

jpv in France

B Smith posted:

No, I must respectfully disagree.  Nah, we're not actually disagreeing about this.  Everything you say in the rest of this paragraph is true and is understood to be true. The "scale" of a drawing (or of a model) is determined solely by the ratio that the drawing (or model) bears to the actual artifact. A 1:48 drawing of an actual 40-foot boxcar will necessarily be 10 inches in length, because at a scale of 1:48, one inch on the drawing is understood to represent 48 inches (4 feet) of the actual object: Therefore 10 inches in the drawing represents 40 feet of the "real" boxcar. If the drawing is reduced in size (or expanded) on a copying machine, then it will no longer bear the same ration to the original and will  no longer be a 1:48 representation of the original object, and measurements taken from the drawing will have to be adjusted accordingly. If the drawing is enlarged 2X, then the ratio will become 1:24 rather than 1:48.

That's what a scale drawing is all about: it represents the original object at a uniformly and consistently reduced size. Mostly true, except it can of course be an increased size, or even an unchanged size, which of course would be 1:1 scale.  Think of architectural plans for a house, which are often drawn at a scale of 1/4" = 1 (1:48) or 1/8" = 1 foot (1:96). The usefulness of such a "scale" drawing is that it communicates necessary information to the contractor without the need for drawings that are as large as the finished house itself. The builder understands that at a scale of 1/4" = 1 foot, a 3/4" inch dimension on the drawing for a certain doorway means that the actual doorway shall be 3 feet wide. The "scale" of the drawing expresses a direct mathematical relationship between the actual physical object and its representation on paper, or on a computer screen. Determining the final or "actual" measurement of the doorway therefore requires a computation or translation, because it is not directly given by the drawing.

Now, in this case it seems to me that what we have is not a 1:1 scale drawing of the locomotive (which would have to be 84'-8" in actual length), but instead a graphic interface presented at some arbitrary scale (the exact scale is not important) which then allows access to the 1:1 data stored in the underlying computer file. In other words, the drawing itself is clearly not at a scale of 1:1, but it does allow direct access to the computer's file of underlying actual dimensions without the need for any further computation or translation.  

We're dealing with a semantic issue.  No arguments on how scale drawings work.  The problem is that when Jpv says "This drawing is in 1/1 scale as you can see with the young lady on the front of the engine and is not at all ready for 3D printing..."  the drawing he is referencing isn't the image posted above but rather the 3d model (which happens to be a drawing) existing in the simulated 3d realm of the Sketchup program.  So what we have is this:

"That's what a scale drawing (your concept of drawing, as in the image posted above, clearly not full size) is all about: it represents the original object (Jpv's concept of drawing, the loco existing in the Sketchup program, at full size) at a uniformly and consistently reduced size."

So you are both talking about drawings, and you are both correct, but you are talking about different drawings.

B Smith posted:

Your final comment about measuring a reduced 11-inch photograph of the original locomotive in which "the bolsters aren't actually a real 43'-6" apart, but rather about 7" apart" only serves demonstrate my point that the photo (or a drawing made from it) is not a 1:1 scale representation of the original object. By definition, a 1:1 scale drawing or other representation of the original artifact must be exactly the same size as the original, otherwise it is not 1:1.

I was never presenting against your point.  I was only trying to explain the distinction between the representation (the image way up at the top) and the original object (the 3d model).  I clearly overthought and overstated my response last night because my last post says essentially the same thing but much more succinctly.

Thanks to Jim (BIG TRAIN), I noticed I had make mistakes on some parts; he gave me the necessary remarks with drawings explanations and I have modified the pieces.

The front short hood (longer), the rear hoods (places of the doors) and the left side walk are now in accordance with the real SD40-2.

Here is the new drawing of the shell:

Some parts have been also moved, for exemple, the tool box under the cab and the filter screen behind the cab. You can compare with the same drawing I posted earlier.

jpv in France

Don't forget the SD40-2 "snoot"?  It had a 118" long short hood if my memory serves.  UP ran them and I'm sure some other roads as well. I think OMI Is on the only one who did a "snoot" version of this loco. 

BH

BH, I have seen these SD40-2 with their looong hood in front! I find that this long hood breaks the appearance of this engine.....matter of personal taste.......

Thanks for your comment! 

jpv in France

Totally understand, it's a matter of preference. It was a string of (5) UP "Snoots", parked in front of the Kearney, NE. station in 1977, that brought me back into model railroading.  I hadn't paid attention to a diesel loco since the mid 60's, as I prefer covered wagons to hood units and high school, college and then marriage took precedence over model trains. 

But, when I saw this magnificent lash-up of SD40-2's, my jaw dropped.  I didn't know a hood unit could look this good.

I had to stop and take some polaroids and still have them. To me, the longer nose just adds to the immensity of the loco and actually balances the loco for me, making it one of the best looking hood units ever. 

BH

In other words: front long hood or front short hood, this engine is always appreciated!

Tomorrow I will talk you about the most important step of this project: the trucks!

It's time to sleep here!!!

jpv in France

I think what you folks are trying to say is that drawing is in the correct proportion to a 1 to 1 version. There is NO way it is 1 to 1.  Another way of saying it is that you could blow up the small drawing  (incrementing all parts at an equal rate)  and it would eventually match a 1 to 1 version in size.

Simon

I know, Simon, but all is matter of words! what is important for me? One drawing gives me the real size of the engine: fine; another one gives me the 1/48 scale, only because I have scaled the model in the software: great; it's all what I need! now, those who will try to blow up the drawing must have good lungs! 

Now, let me show you the trucks:

At the beginning of the design, I didn't want to have motor (s) inside the shell; all the area would be for electronics, wires, etc...

So the only solution was to have motors inside the trucks; the design was then made with this idea; I opted for nose hung motors like the real engine. I had noticed that NWSL made the Magic Carpet in this way but I had no drawings; I mailed  for informations and they answer me that they hadn't any drawings for this item..........(?)! they answer me that I could buy an item to translate it into 3D drawings! well, but if this model didn't fit? What do I do with?

I give up this solution and create my own hung nose motor!

Here is the 3D drawing of the whole truck:

The middle axle is dummy (not enough place under the bolster!); the two other axles are motorized.

I have chosen a FAULHABER gearmotor (compatible DCC) with additional gears outside (speed reduction ratio 1/3) on the advice of a friend.

Here is the truck chassis; the bolster will be fixed on with 4 bolts and suspension springs.

Detail for the hung nose motor:

As you can see the journal boxes have also suspension springs.

It's a general idea , but I hope it's enough clear! 

If you have any question, I will answer with pleasure.

jpv in France

It's a long time since I didn't speak about this engine! neccessary adjustments and other things.....and my "3Dguy" was very busy with big orders to deliver; anyway..

Here are the first parts in 3D I received some days ago; I added details which can't be printed because too small: nuts, hinges and anti slip strips (Ho scale pattern, O scale is too big):

And the chassis:

In 2 parts which fit together with the help of the little tab (bottom left). All the holes are for the fittings of the shell components  on the chassis with screws and nuts (glued inside holes planned in the components);

like this (assembly test):

Hope to advance in this project faster next year!

Thanks again to Jim for his help and advices!

jpv in France

Looking great as always!  Off to a great start!  Looking forward to seeing all the progress on this build!

Beautiful work as always. You are a prodigy man. Pure genius 

So just to get a handle on this stuff:

Let's say that I wanted two of those Flexicoil truck castings - bare, no journals, springs, brake cylinders, etc., just the castings.

What would it cost me to have somebody in the business print me a pair in structurally usable plastic?

How about the same in metal?

Is this stuff actually commercially viable for model trains, or am I looking at a grand for the metal parts?

Bob,

Do you specifically mean the EMD HTC-1 sideframe as 3d modeled above?  Or are you just asking generically?

If you specifically want the HTC-1 casting, then the first step is to look for CLW castings for this truck style.  Or see if you can access existing molds and have new ones cast.  The CLW castings are actually pretty good, and scale out pretty accurately.  And they are designed to accept springing and journals like the P&D Blomberg trucks.

If you're just asking in general, then the answers are a little different.  First off, I'd argue that there is no "structurally usable plastic" available for printing at this time that will yield the type of resolution and detail we seek as model builders.  Of course, you're a doorstop sand cast kind of guy, so maybe your goals are different.  But any plastic that will print with any level of resolution and detail probably won't be structurally sound.  For instance, the Frosted Ultra or Extreme Detail plastic from Shapeways will print as low as 16 microns per layer I believe.  But it is brittle, especially in the types of cross-sections we are dealing with.  Nylon is much stronger, but it can't print as finely as the frosted detail plastics.  There are other places that can produce better prints, for more money usually, but typically the goal is take a print and use it for a master for resin or metal casting.

As far as price is concerned, I'll use this EMD Flexicoil sideframe casting as an example.

From Shapeways, it would cost about $25-30 to print this model in Frosted Extreme Detail.  It would be incrementally more expensive to print a 3-axle truck.  From a printing or rapid prototyping house with better printers (or a custom jeweler even), I'm guessing it would cost about twice as much for the same print.  I need to get some quotes to know for sure.  You could also have masters machined in wax as opposed to printed in wax, using a CNC mill running off the 3d model.  The mill will produce better resolution, but there are the same issues with undercuts as with any machining process, so the master will have to be designed differently.  I think the milling method will also cost more, based on my limited contact with a custom jeweler who was willing to do the work for me.

Once you have the masters in hand, then casting in metal or resin doesn't cost any more than it would for anything else.  What has it cost to have trucks cast for you?  I know you've had it done for some of your models.  Of course, you could potentially offset your costs by making a mold and casting extras for sale, and the mold won't be free.  But for a one-off casting, you could probably get better detail from a direct burnout.  Of course, if you do direct burnout, you need two masters for the pair of trucks, so double your printing costs accordingly.

I need to get a quote for casting these trucks in brass.  I'm also looking at casting in resin instead since these are intended to fit to the Atlas sw8/9 gearbox.  They won't actually be bearing, just along for the ride.  I was looking at resin casting the brake rigging anyway to avoid electrical isolation issues.

By the way, Shapeways quotes a price of $133 to "print" this sideframe in raw brass.  However, the 3d model currently doesn't pass the test for minimum wall thicknesses.  But the reality is that Shapeways "printing" brass parts is actually just direct burnout lost wax casting.  They print a wax master, cover it in plaster, burn out the wax, cast the brass into the plaster mold, then break away the plaster.  I'm pretty sure you already know somebody that can do that for you.

Hope that helps,

Jim

Very nice and informative thread!

Not very knowledgeable on the UP SD40-2 options but I take it they had On-Board Rail greasers as depicted in this square cabinet next to the traction Motor blower duct:

That's the electrical cabinet air filter box (ECAFB).

maybe this is an option for your powering the SD.

http://www.abcgears.co.uk/html/traction_motors.html

an other option is make them yourself, i did something simular

Cor

Thanks to all for your comments !!!!!!!!!!!! very appreciated! Lot of work to do on this engine..............!! and now that the building of the SD40 has begun, I can keep on working (3Ddrawings) on the next project! but shhhh! it's a secret! but I'm sure you would like to know !

MOONLICHT------>Thanks for your suggestion; I know abcgears; their solution is good but don't fit inside my truck frame: a little bit too large! NWSL' magic carpet seems good also but they are incapable of sending me a drawing; I would translate it in 3D and make an attempt! it's unthinkable!

The one you do looks very fine and accurate! 

See you soon, gentlemen!

jpv in France

Jim - that is precisely what I was asking.  Shapeways may be slightly high, but not much, for burnout.

Yes, I have used CLW Flexicoils and PA Trucks.  Cannot be beat for what I do.  I have just about everything a kid could want in O Scale except for the sand cast Erie-built, and for that I await only the sides.

Costs for rubber molds have gone up considerably, but I have a dentist friend who can make cold molds, and I have successfully sent candle waxes to the foundry!

Hello all,

The 3D parts are long to come due to some repetitive failures when printing; I received yesterday some of them and could make the assembly of the hood just behind the cab:

I hope that my "3D guy" and me will be more lucky now!

If the weather is not too cold tomorrow, I will give a coat of primer.......but we are in winter..............

Parts to come: the two chassis ends platforms with the steps and the fuel tank.

jpv in France

Looking great!  I can't wait to see the other parts added.

Many thanks, Jim!

'Trying not to be too long to add parts!!!

jp

Fantastic work. Very entertaining to see your progress 

Thanks you very much, Eric!

jpv 

Really exciting work and fun to watch progress!

Thank you, Mike

See you soon

jpv in France

Received today (earlier than I thought) some parts: end platforms and stairs.

The white pieces are made with PLA filaments, the stairs are made of resin, more accurate for the stairs round pattern (quite impossible to do with PLA). And these parts are not big! (rule in mm).

Now the assembly platform/stairs:

and fitted on the chassis:

These white parts make the photos ugly!  Sorry, gentlemen!

Tomorrow, the other end side and photos of the fuel tank! I have to sand it before.

jpv in France