I may get a little long-winded here, and at first I will be talking about some general 3d printing stuff, but very shortly I will be getting to some specifics of the Orville PM44 with some build along screenshots of my work flow, since there seems to have been some interest in going from concept and art to 3d printed thing.
So, for starters, I will have a much different end goal than say, RollerGirl_TashaSkar's technique. Not that there is anything wrong with the two halves technique. I just know that I can't keep people from opening doors and windows and letting cold air warp my parts. Tasha's way means you don't have to worry about bed adhesion so much, and you will just about never ever need to use support material. My way means that I can keep the layer lines parallel to some of the major contours as opposed to perpendicular making some parts prettier, but structurally weaker, but really this boils down to building for your printer. If you look at the inside of the top parts of my TNG Medical scanner, I slope the inside pretty significantly because my printer does overhangs really well, but is terribad at bridging. This lets me minimize how much support material I need to get a good bridge that doesn't make me cry. Also, to explain "overhang" and "bridge" in this context, think of the capital letters Y and H. Y shows overhang. The arms stick out to the sides, but they come out gently, so they can get a little support from the layer below them. The horizontal line on the H is a bridge. The edges are supported, but the part in the middle has nothing to hold it up. Some printers do one or the other well, some do both well, some do neither well.
So, I want to make my parts taller than they are wide, I'd rather have overhangs than bridges, and if I can, I want to limit my seam lines to where there are seems on the prop. There aren't a lot of seams along the length of the thing. The "barrel" is pretty obviously cut in sort of half along a curve, then there is a pretty obvious seam between the swoopy barrel and the grip, and then a mess of seams around the grip and the panel with the buttons. Despite the fact that I am going to start over, this is the same plan I had with my first attempt at the design, so things are pretty similar up to this point. Originally I had the barrel cowl as two parts, the inner barrel as a third, the grip and trigger as a fourth, and the control panel as a fifth.
Now, thanks to the new pictures ESPECIALLY of the grip, I am rethinking the need to have the control part separate, or to use it as part of the connection to the barrel like I did on my first run. I may want it separate, but I will probably not make it integral to that joint based on this image here.
Now, the next trick is deciding scale. Without a ruler or an object of a known size in any of these pictures, that means we are left to our own devices for scale. For a pistol like this, I scale it to my hand, since I'm making it for me. How to scale it to your hand? My favorite trick comes from Bill at Punished Props. Take a reference photo with an actual pistol/airsoft/BB gun/toy gun that fits in your hand well.
Failing that, measure a line across your knuckles, and add about an inch for comfort and organic shaping. I have about 3 inches across my knuckles, so four inches from trigger to where my pinky will rest or there abouts.
Now that we have a rough idea of scale, and how we want to break up the project, it's time to actually go into a program and start making shapes! There are a lot of different programs to choose from, and a lot of different workflows available. In my time with this forum I've seen people using Z-Brush, an excellent digital sculpting program though it does have a bit of a learning curve and you really ought to use a pen tablet like a WACOM instead of a mouse, you could use more traditional modeling programs like Maya or Modo, you could use any CAD program... or something like Blender that is more of a jack of all trades so you can have a lot of different work flows available to you... really, it's up to how you like to work and what sort of budget you have for the software.
My background in this sort of thing is more from the animation side of things, so I have a lot of bad habits that don't translate well to 3d printing, and to help curtail them, I try and work in the CAD software where the tools that let me get away with those bad habits don't exist. Since (at least for the time being) Fusion 360 is free for hobyists, that's what I will be using.
Step one: Set up the work space.
I make a new project, and there are a few things I do to make sure my brain and the program are working together. Z is up/down for me, X is left/right, and Y is forward/backward. That's how it worked for me in math class, that's how my CG graphics package works, that's how my 3d printer is set up, and it means I don't have to translate anything. Next, I'm going to load in at least one reference image. I'm using the side on photos to start, as the most important part to me is the silhouette of the prop. Once I bring it in, it's time to scale the image. In F360, this is done by right clicking on the canvas in the browser, choosing calibrate, and setting up the grip to be the 101.6 mm I decided on earlier. Now, I default to Metric, but the program is smart enough that if I put "4 in" in that box, it will do imperial as well.
So we get a few pictures from similar angles, and we can stack them up a little...
Now, this does create some visual noise but we can kind of become aware of the angles of the images and get a better feel for the shape. You will notice I have the XY plane coming across at the seam between the barrel and the handle, and the origin lines up on the split between the barrel halves. This is done on purpose, so when I export those parts as STL files, they will be in the correct orientation for the printer.
I'm going to use these images to extract the curve for the barrel, and I am making an assumption here. My assumption is "if you stretch out that curve, it will eventually make a circle." I'm going to add a sketch, and I am going to name it early and often! It will be in the YZ plane, and I will use the 3 point circle tool since that is a small arc of the circle, and i have no idea where the center is. My first point is at the origin, my second point is at the "dangerous" end along the seam, then I fiddle around to fit the third point so the circle fits the arc.
With the middle of the arc defined, we can now easily measure the top and the bottom halves. I had initially thought those were symmetrical, but in light of the recent images I am pretty sure I was wrong. Using the circle we've traced, I'm now going to use the offset tool to make concentric circles to measure the distance to the top and bottom of the barrel shape. Both my images agree the top is about 30 mm from the seam to the top, and between 35 and 45 mm from the seam to the bottom. I split the difference and called it 40 mm.
Now that we have the height, we need to figure out the width and the profile. To me, it looks somewhere between an oval and an ellipse. Ovals have too flat of sides, and ellipses come to more of a point... so I'll play around trying to get a shape I'm happy with. After doodling a bit, I think I like this profile for this attempt:
I am then going to use the sweep tool to spin it around that first circle we got for the barrel and... We realize I made a mistake!
After all my careful measurements, why did this happen? Because I made the profile sketch in the wrong plane. I made it around the origin. It should have been placed at an angle so it makes a right angle with the circle I'm using for the sweep tool.
We have a lovely hoop, but that's not exactly what we want. We need to trim down some of the tube for a barrel. Cut in another quick line on the barrel arc sketch, and now we can extrude the parts of the hoop we want to cut off. Select the negative space, basically, then extrude it symmetrically, and make sure the operation is cut. Then all we are left with is this:
The next thing I want to do is cut in the "trigger area" and the place where the screen goes. I am going to make another sketch in the same plane as the Barrel Arc for these, and again, extrude them to cut away parts of the barrel object.
Offsetting the circle for the trigger guard gets us the raised bit that flows into the trigger. I use a few fillets to round off the shape of it as well as the edge for the display, and then using the shell command, I make it hollow like a tube. How thick should it be? Well, cutting the barrel in half you get about 4 equal slices. The outer cowling, the stud detail, the inner barrel wall and then the gap. In my version half the barrel is 20 mm, so divided by 4 gives us 5 mm per layer.
Next on my to do list is to split the piece, putting the seam in. So you know, I am about to use what appears to be a "magic number." This number seems to work well for creating the gaps between parts on my printer with how I've tuned it. I came to this magic number by printing a series of parts that are supposed to fit together, one box and cylinder inside another. I made the inner part smaller in .1 mm increments, until I got two parts that would fit together with friction, but could still be pulled apart. The gap that worked the best on my machine was .2 mm. This is half the width of my nozzle. At some point I will try a different nozzle to see if that affects it, or is just coincidence.
Time to bring back my barrel arc sketch, and add a few more offsets to the first line. That curve is probably the most important thing in this whole project. I am going to offset twice, once by a positive .1 mm and once by a negative. Before I do the cut, just in case I want to keep a copy of the barrel that is whole. this is not obvious immediately in Fusion 360. To do this, right click on the object in the browser and select Move/Copy.
The obvious and yet easy to miss part is you have to put the check mark next to "Create Copy." It took a while to realize that box was there. Learn from my mistakes, it will save you so much time!
You will then have to hide the original by clicking the lightbulb next to the object. Otherwise it will want to merge the copy with the original basically doing nothing. Upside, it won't let you hit the OK button on the copy dialog if that is the case. I then renamed one of the copys Whole_Barrel, so I can have it if I need it later.
Hide the barrel, zoom in REALLY CLOSE and select the two little slivers so we can split the barrel. Extrude and cut the wafer thin slit, rename the new parts to stay organized, and on we go!
But On we go on another day! I hope this is helpful to some people, and please, let me know if you have questions, if you think I'm going into too much detail, if I'm not going into enough detail...
Just so folks know, the next couple of targets will be creating a way for the two halves of the barrel to be glued together, and to make sure they line themselves up as much as possible, and also have a way to register both these parts to the inner barrel detail.
As you can imagine, taking the time to document the steps like this seriously slows down my workflow, but I am enthusiastic about this project, and I hope to have at least the registration done by tomorrow night, or Monday at the latest.