My current weekend Project: FFVII - Squalls gunblade

Created by TsunamiJuan on Sept. 4, 2014, 3:20 p.m.
  • I feel comfortable sharing this project now that its got a good deal of steam going behind it.

    I am working on making a functional Replica Squalls Gunblade. Big stress on functional. As the original design was never meant to be a real item. There is no exit point for the barrel, there is very little room for functionality on both the Hammer and trigger. Not to mention the advance mechanism for the Cylinder. So I am having to do a large amount of engineering while trying to keep it true to the original concept. Though there is some things like the squareness of it, that I feel need to be softened a bit. So that its comfortable to hold.

    here is one of the more accurate examples of Squalls blade. Its the top one. Its a 3" casting in plastic for the official action figures. I bought both of them 15 years ago cause I have had a fascination with them. In fact squalls blade was the first object I ever created a 3D Model of.

    Its basically a mirror of itself on the other side. The only difference is the addition of what appears to be a rotating knob assembly, to remove the tension and lock on the cylinder of the gun. That side is also the side that the Cylinder and swing arm assembly swing off of. At least according to the master arms replica that was licensed and produced w/squarenix . I am trying to get my hands on that version but I have found reference photos of it to pretty much fill in the blanks. They had to do quite a bit of modification to make it semi functional (hammer, trigger, and swing arm with cylinder).

    I figured I would start with a Inside out approach. after my interested was peaked again to do this. So I first needed to figure out the ammo, and cylinder size. Once I found that I could scale everything else from there. Which worked out well since I knew the case for the gunblade is roughly the size of a electric bass guitar case. (I got guitars and cases galore around here).

    I started off with making a 3D model of 0.50acp ammo in solidworks. Off of what i later found to be a somewhat incorrect set of dimensions. Not a big deal since I was already planning on using a moon clip with the shells at this point. Which turned out was also included in the master arms replica by chance.

    I did a 3D print of the shells and found them to be to small scale wise for what I felt was needed. They where just to small. They didn't have that Imposing feel, or bulk. So i doubled the scale of it and did another print.

    Success, the result is double the scale of the 50acp on the right. Leaving it with a slug that is roughly 25.7mm in size, and a casing that is almost 29mm in diameter. Now we have an ammo size. Next we need a moon clip, and a cylinder. Back to solid works to mock up a quick cylinder with approximate clearances. Can't forget the indents on the sides between the shells as well (i am sure these have a name but I don't know it). A moon clip was mocked up, and printed and checked for fit. Then adjusted a few times. While at the same time making changes to the Cylinder clearances. I was about out of time that day. So I tossed the 3d file for the cylinder to the printer (happily in another room) and let it settle in for a 7.5hour print. Of the roughly 95mm diamater cylinder at this point. with a height of 85mm.

    The next morning I pulled the print off the printer and took these shots. Yeah it looks a little like the attack of a nerf gun right now. But nobody going to mistake it for a real gun yet.

    This is about the first day or two of work coming together here. Its quite a bit further than that right now. But I am out of time to write up and show the next pieces and progression. So I will save that for later.

    Hope people enjoy the pictures and the geekery/fun i am having as well.

    (currently I am 3 weeks into the build so there is quite a bit more to put up 9/4/2014)

  • Part 2

    Once I had a feel for the scale of the cylinder and ammunition together. I knew I still needed to make a bunch of changes to what I had created so far. Which was a great learning chance in solidworks. So I could fairly easily go back and resize and rebuild components as I went.

    I did some initial scaling of the design, around what I had done so far.

    This was a super long and involved process. I spent hours with a calculator, a sketch pad, and a pair of dial calipers. I was also braining storming during the process trying to figure out how I was going to split the design for printing. As well as how much room I actually had for the mechanics of the gunblade.

    I then decided to take a picture of the gunblade model I had, on a flat plane. Then compare it to my scaling in solidworks. As it turned off I was close in a lot of places. But extremely far in others. This bummed me out quite a bit. However I decided to just move on and conquer more of the mechanics of the build. Since I really didn’t understand them well enough, or at least they were not clicking fully in my head.

    I decided to start with the first part of the cylinder advance mechanism. I knew I needed a shaft that went through the middle of the cylinder. Then through the swim arm. I also knew that said shaft/spindle, extended a good distance past the swing arm. This is very much a key aesthetic feature of the design as much as function.

    As I don’t own a revolver. Have never handled one, or fired one. I needed to do a good deal of research on gun designs. In the process I learned that cylinder moves on the spindle in a back and forth motion as well. In order to assist in the ejection of spent cartridges/rounds. This was not something I had previously planned for. Thinking that having a moon clip in the design was mostly going to take care of this.

    There was also a matter of how I was going to turn the cylinder with the double action from the trigger. I was thinking that I would use a finger that locks on the fingers on part of the cylinder. However there are a large number of possible designs one can use when it comes to doing this. I ended up setting on one directional gear/clutch style design. So that the finger would push in one direction then slip back correctly on the return.

    It was another great solidworks learning experience to make this. I ended up making it one triangle at a time. Then extruding each one. Then making a cut across the surface from the side. Once that was done, I mirrored the triangle in a circular pattern. So that I had a total of 6 segments. Then added a self-centering/indexing pin in the center of the part. So that as I swing the cylinder back into place it should be able to center itself. It should also deal with the compression from the locking mechanism behind it.

    From here designed the shaft having a hex head on part of it then transitioning into round shaft. So assist in the ejection function.

    Now that I had part of this designed I needed part go on the gunblade body itself. Also to make sure my Tolerances where correct. So I had printed a couple of samples of the plates. I took a copy of the same part and cut it so it was circular for the other side. So I didn’t have to worry have double work.

    Once I had worked out how I was going to mate and rotate the cylinder I was itching to get another print done. However I wanted to update the Cylinder design to include a mate for shaft. I also wanted to fix the profile flaws I had made on the cylinder to begin with. The indents as I found where not spherical at the end. Which Meant I needed to go from a fillet at the end to actually making a lofted profile to cut into the cylinder. This took a good deal of learning in solid works. I also added the indentations for the other set of rotation locks. Which took more thinking outside my current knowledge.

    The trick to this was thinking of how I would do this if I was doing it by hand at a tool mill. However I am not a machinist. I do not own a vertical mill or knee mill, or have ever used one. Luckily I have had a fascination with buying one for years. So I knew that you would likely use an indexer for cutting gear teeth. Or other similar spherical changes. So you could go off the correct locations of degrees. To be able to work on a squared plane of reference. By figuring this out I ended up combining a small arc that was extruded and cut in. Then an elliptical path so that I Would not cut through the other side of the profile, and still have a tiny lip so that the lock had something to hold onto. This brought me pretty much in line with how an actual revolver cylinder is designed.

    Now I was finally ready to do a new set of prints. So it was once again time for another 8 hours or so of printing the cylinder. Accompanied by an hour or so of time to print the shaft, and 15-20m or so for the updated moonclip (which is floating above the cylinder in the above images.

    So here are the results from the print.

    Still more to come, I am about 3 weeks behind of where I am currently at in the project. When it comes to writing this up and taking pictures. (posted 9/11/2014)

  • @MantaBase: Solidworks 2014, If you have an student ID you can often get a legit copy for around 90-130$ for a one year license. Which is honestly better than paying the money for any of the autodesk 123D suite of products as they are buggy enough still that they can prevent you from using them reliably. I was using 123 Design some, but when you start generating sketches with complex fillets on them, they software will crash without warning.

  • Sadly I hit a nasty wall again this weekend, the mechanics of this project are warping my brain. But I figured I would toss some shots in here of some of the test prints assembled to this point.

    A simple series of test prints, Working out tolerances and the dimensions of the swing arm and body. The series of 3 small registration rods, was only for test purposes. The orange frame was to test the cylinder rod and alignment with the advance mechanism.

    Pictures on misc prints I have made to this point. Most of the exterior of the assembly. The weekends big prints. The two handle tests. The blue one is externally dimensional correct. However the internal structure of it is way off. The Orange is internally closer to what I wanted as an example for a lot of this build. However again its not quite correct for a handle. Both are 4 piece prints for the exterior. However only the orange one has registration keying, lips, and a small block in the center that friction fits and holds the top and the bottom together snugly.

    This has definitely been a huge learning process. With each stage getting more and more complex.

  • I didn't do any work on this for a few weeks while tweaking printers. I upgraded the bed and XYZ axis sizes. It took some tweaking to get everything happy again. Not to mention I got sucked into making parts to improve the cable management so that they wouldn't hit and catch on parts on the table. Which helped me learn how to design and print small parts. I also switched slicing software and started printing ABS.

    I hit a wall yet again with my trigger mechanism design. I was going to use a ratchet and pawl on an axle. Which I designed a spring mechanism for and expanded on the clutch and locking systems for. However there just isn't enough room inside the body for it. I could change the body. Though if i change the dimensions of the exterior it throws off the work i spent scaling everything. So that it was as true to form as possible.

    Spring and ratchet+pawl axial assembly, The spring goes in the back section of the ratchet and paw setup. The Pawl also acts as a Stop so that the axle doesn't eject out of the body while the cylinder is removed. To bad I can't use it now...

    So on and off between other stuff today I have been modeling and printing and revising a new trigger assembly. Its got some stress issues so far. But with using ABS for these small parts instead of PLA it seems to make a huge difference. Though it might come down to machining them out of steel or aluminium for the final product. Since I don't think Dissassembly to repair it will be an option. At least not at this point in the design.

    I've been holding off on completing the blade portion of the model. Since it throws off how I compare it to the current scaled images I am using to track if stuff is scaled correctly or not. Honestly the blade is by far the easiest part. Once the trigger assembly is finished I need to choose a accessories rail system to use. I was thinking a picatinny rail. Though I am not sure if that is going to fit in the area I have to work with. I had discovered quite recently that the top of the gun assembly is intended to have a pair of iron sights on it. These sights line up with where the top portion of the blade and barrel come back into view. Before staying mostly straight till the small dip before the tip. (which might be were the possible exit point for the rounds is).

    I also need to find a way to break up the design some probably with chamfers and possibly some Knurling. So that it doesn't look like a shitty 3d render. It needs to have an air of functionality to it IMO.

    More to come :)

  • Glad to see an update on this project :)

  • This is so cool! I love how you print everything. What printer do you use? I am about to build one for my home (first trying to fix electrical issues with the one I'm building for work which I gladly use as a test dummy, lol) and people often ask me what the hell I will print. I should point them to this topic, this is one of the more elaborate prints I've seen.

    How will you paint it?

  • @General Desire I am using a printrbot metal simple, W/heated aluminium bed upgrade for the ABS, which also did most of the PLA printing till recently. Then i have a second one that has been upgraded with the Matrix Precision v2 bed upgrade kit, and firmware. The differance in flatness between the Printrbot bed and the Matrix bed is huge. The machining quality and precision is noticeable. I just need to get a 8x8 heating element to be able to print ABS on the larger setup.

    I originally thought when I bought a 3d printer I was going to use it mainly to produce prototype parts for my camera setup, and some miniature scenes. But I am constantly finding new uses and problems to tackle with it. Like I designed and printed one piece roller bearings for my filament spools. I also have printed replacement parts tools that had damaged plastic casings or parts. Honestly the imaginations the limit most of the time.

    As for painting I am thinking I will go with a medium build airbrush-able primer. A high build primer would likely throw the dimensions off to much and cause serious alignment issues. All I really need to do in most places is cover the ribbing left from the printing process. I am not running into much ooze and over extrusion anymore. So I won't need to worry about bumps from that as much. Even though the size is large I will likely airbrush the paint on as well. Since It will give me the ability to put down a nice thin layer. It will get a degree of weathering as well. Since i am going to try to emulate a brushed metal finish. The kinda of finish you would expect to find on a decent knife blade. Which will require me to intentionally scratch the surface and then polish it. Same needs to be done in a similar manner for the blade edge. I want it look as if its been ground by a knife makers belt sander to a certain degree.

    I will likely spend just as much time painting and finishing it as I have building it up to that point.

  • Happy Holidays,

    So its been a while since I have posted anything, been doing a lot of wracking of my brain on the trigger mechanism. I am semi reluctant to post some of that stuff on here to avoid in questionable legal issues, or morality concerns. So The likely hood of showing pictures of the sub assemblies of the mechanism are low.

    WARNING - I am going to discuss firearms some in the next bit. If your offended by this consider yourself warned. Also I don't condone gun violence in any form. But I do advocate proper training and safe handling of weapons if your so inclined. My views not the sites.

    So I have been going through gun schematics getting a better feeling for how others designed such assemblies in the past. I happened upon a schematics for a Colt Python. Which seems to be very close to fitting in the size constraints i am working in. I have also moved away from using 3d printed springs. They are just to large and unreliable for what i am putting them under. Not to mention difficult to work with. So I have moved to using actual gun springs. A few of the gun smithing sites out there sell spring kits. Of varying sizes and strengths, which are ideal for prototyping with. They also sell spring steel rods and. I am sure I could talk to a gunsmith and get what i am looking for. However having a complete kit as my disposal for the relatively low cost of 30$ a kit seemed like it was worth it.

    I have also moved on to version two of the prototype. Which incorporates the trigger guard into the body at the correct scale. It also Involved the hollowing of the handle and some other major thickness changes. Resulting in a much lighter and more comfortable body. I also have incorporated a side plate, Similar to that on the Colt Python. To accommodate easy access to the double action assembly. Similar to the Python It will be held with a screw or two, and a friction fit. This both helps with the structure of the gunblade, and will be more aesthetically pleasing then the original plan of having the body split like a toy gun into left and right halves.

    I did a test print of the Blade portion this weekend as well. I wanted to see how it was going to feel scale wise. This first version of the blade incorporates a barrel through the blade. I didn't think it would work without throwing the hole design off but it actually complemented the scale quite well. It comes out just above the notch on the end of the blade. If it was a real weapon this would be a horrible horrible thing with a high tendency to cause the round to curve or even hit the blade, But its not so nuff said. The blade thickness also works out wonderfully. Its just the right level of ridiculous, its not to fat and its not to thin. Its roughly 36.5mm at its thickest.

    So for comparison, The geek Chic Hero Sword, a yard stick, Blade Prototype 1 + Gunbody Prototype 1, one unloaded Gunblade shell, and the gunbody prototype V2. The gunbody prototype 1, is shorter than it should be, so the actual design is longer still. by another few inches.

    More and more changes still to come, but its quite exciting to have a good feel for it in my hands. I am so happy that the scaling to my person has worked out so well. (I am 6'3" ) . It doesn't look tiny and small in my hands. It looks to have the expect proportions one would expect in relation to the game.

    Happy Holidays and a Happy incoming new year to everyone.

  • Love this project, and also the crazy colors haha :)

  • This is an awesome project. Squall's in VIII, though, yeah?

  • @tmee: Yeah messed up the title and it wouldn't let me go back and change it which is fine. Those in the know will have a good laugh :D

  • I actually had a decent amount of time to throw into this the last few weeks, on the actual design department. I've spent some time working on where the blade meets the body. Trying to smooth out the way it looks and fix a few small scale issues that have popped up.

    It turned out the blade is actually about an inch and a half to long currently, and the cut where the barrel exits over the tip, is to short. While i like the way it looks I will probably change this somewhat to be more closely inline with the scale. The blade is already more slanted than than the action figure model. Which can be explained as the slant was so minimal they didnt bother or was poorly molded. That or they just never took into account functionality for barrel exit so it was never there in the first place. Either makes sense or both.

    I also started doing some serious work on the trigger. Finally making enough success to move forward. The trigger now advances the cylinder. Its not properly indexed at this point but its a huge leap forward from where I was. It also fits properly within the body of the gunblade. Todays work is to move on to moving the rotation axis of the trigger forward, So i work on the finger that locks and holds the cylinder from rotating until is released by the trigger for advancement. You can see pencil marks in the picture for about where stuff needs to move.

    Once that part is worked out. I will then move on to the hammer and rebound arm. There is also the challenge of getting the hammer installed into the assembly in such a way that it can be removed for service/replacement but not ruin the scale/aesthetic. Its possible that I might slide it in from the outside through the slot around the firing pin. But its still kind of an unkown at this point till i get that far.

    I have also ran into odd printing problems. Mostly cause I tend to print stuff so quickly and its been so cold here lately that its a blanced between getting the bed too hot, and to cold. Also since i tend to print at high speed so i can get stuff in hand somewhat quickly I have found that it leads to some inaccuracy in the printing of tall items like the blade where stuff has become shift partly in the direction of the print head. At this point it does not matter, but It might mean that the final prints take much longer than expected. Or I divide the parts into much smaller pieces for printing and add indexes, Then spend more time gluing. It currently takes me about 4 - 6 hours to print a revision to blade body at around 3600mm/min. This is fairly heavily optimised by Simplfy3d it would probably take me longer and require me to print at 1200-2400 in slic3r while having severe accuracy problems.

  • @TsunamiJuan: Looks awesome, good job. Just out of interest, do you use Repetier as a host for printing, or e.g. Pronterface?

  • @General Desire: The host I use for printing is OctoPrint, as my printers on plugged into the physical network in a different room. So i can print all night and not annoy people with them. I should be able to print right from Simplify3d as well. But I have never tried it. I generally just save the gcode then move it over to octoprint, and handle the print from there. It also allows me to check it remotely If i so choose too.


  • @TsunamiJuan: Ah right, that's what I have heard Will and Jeremy talk about as well I believe. I was just asking cause I use Repetier for the Cartesian printer I already built, but the Delta I am finishing up right now moved terribly slow on that one. I like the tinkering and troubleshooting involved with 3D printing but there seems to be a certain degree of randomness whether something works or not. Anyway, I switched to Marlin and Pronterface now, which is a bit less GUI-friend, but I think it's a good thing as I get to understand the underlying code better.

    Printing at night.. do you have a smoke sensor/fire extinguisher installed? :D

    *edit*

    Aha, could be interesting to try and get OctoPrint to work on an Intel Edison, something I've been meaning to play with. First let's finish this Delta...