Jump to content

RideHPD

TU Member
  • Posts

    11
  • Joined

  • Last visited

Everything posted by RideHPD

  1. No problem, this is the side of engineering I actually enjoy, and it's much more fun when it's not your headache to solve. In a root cause analysis you try to isolate variables you think are responsible so you can eliminate them or identify their role in the failure. In this case there are way too many variables and not enough data points to identify a single area of focus. Was the one that cracked in 36hrs subjected to heating in your car? There's a couple things that stand out, namely the heating cycles in your car and whatever loading is occurring while fishing. Have any of these cracked that have just sat around, without being thermally or mechanically loaded in any way? That's the first thing I'd want to know; is it occurring just from the processing itself? It's hard to really gauge what's going on without having the material in my hands to see what their relative stiffnesses are, but it sounds like there's some residual stress being introduced from heating or loading that are being released during cracking, and my first bet is that it's thermal. Pour a batch of 6 lures, all identical. Get a cheap toaster oven (do not put plastic in your kitchen oven) and soak 3 of them at 150F-ish for a couple hours, then once they've cooled coat all of them in epoxy. Then leave them your car as normal. I'm thinking the foam is curing in your car at higher temp while expanded and that's what's introducing the residual stress (upon cooling). Then repeatedly going between hot car-cold water, etc. is fatiguing that interfacial bond.
  2. So, some things as they come to mind: This lure that broke in 36hrs, was it fished at all, or just resting that entire time? Definitely possible that different areas cured at different rates, and not just radially, that's standard. In the image, were lures (2) and (4) fished at all? Have you tried bending an uncracked one to see if it would crack? Bending the cracked one may or may not yield any usable info, as that crack formation would relieve a lot of the stress that was built up prior. Overall I'm getting the gist that these lures are failing without being fished, or otherwise loaded, and that those cracks may just be initiating from the largest defect in the interface. At this stage I'd start recording data on every lure. Date, time, temp, humidity, relative pressure, what resin, how much mixed, etc. and record when they fail and how, including making some in each batch to just rest and watch for cracks, some to bend, etc. The more you can record, the better you can fall back on the numbers, as it sounds like you may also be seeing process changes that could be due to environmental/seasonal effects. Room temp cure thermosets are pretty sensitive to these things, so you could just be outside of the use envelope or something like that. If the flex epoxy isn't expensive it may be worth applying it to a sample without a turner just to see what happens as a test article. Might find that it doesn't fix anything and save you from buying the equipment. Best thing I learned as an engineer is that it's way faster and better to just get up and try something than burn time behind the desk researching (unless you have a really, really specific problem and you've pretty much proven what the solution is).
  3. For every 10g of mixed Alumifoam add 2.2g of microballoons. (per 1 mL of mixture, 0.1 mL microballoons, 0.9 mL resin; 1.2 g/mL for resin, 2.35 g/mL for balloons, 0.1*2.35=0.235, 0.9*1.2=1.08, 0.235/1.08=0.218 g microballoons/g mixed resin) Overall I would tend to agree with Mark that if the uncoated lures didn't give you problems a flexible epoxy will probably solve most of your problems. That said... Looking at the images, particularly the cracks in lures (2) and (4), and the right-side crack of (1), I'm seeing mostly signs of failure from flexural loading. (2) shows chipping in the coating right at 12 o'clock and then a straight crack down, while (4) shows a chipped coating at 6 o'clock and then crack propagating the rest of the way. Those are pretty common crack initiation site features, which would suggest those surfaces were in tension prior to failure. The right-hand crack of (1) appears to be pretty flat and straight down below the wire and rough at an approximately 45deg angle to the wire, which would suggest tensile and shear failure, respectively. Put together, that also looks like a flexural failure. That makes me less inclined to believe it's the result of a cure shrinkage/thermal dialation issue and more mechanical. Guessing the epoxy is too stiff and is working its way into the pore structure of the foam to an appreciable amount, which is causing the foam to fail completely, something along those lines, though I'm not fully convinced of all of that for things I'll save for now, more off the cuff speculation. So yeah, maybe try a flexible epoxy and see how they behave.
  4. Can you post pictures, including close-ups of the foam/coating interface as well as the foam/wire interface if possible? That will give a lot of clues as to the force distribution prior to the rupture. Definitely sounds like residual cure shrinkage stress issues, but odd that the coated lures cracked, as I wouldn't expect the foam to be able to build up enough stress to cause the epoxy skin to rupture. Foams do experience a greater degree of thermal expansion/contraction, but they're also much softer/less stiff, so they usually just tear themselves apart due to thermal gradients in the foam during the cure process. Working guess would be that the (presumably) stiff epoxy was well bonded to the foam, preventing the foam from shrinking radially, which would cause the foam to also want to contract axially, which led to the circumferential cracking as the weaker foam failed before the stronger foam/epoxy interface did (or foam/wire but that still eludes me). You see this kind of failure frequently, but usually with very stiff, very brittle ceramics. Another useful set of data points would be to measure the length and depth of the mold cavity, as well as the length and diameter of the uncoated and coated foam bodies. That would help identify where the shrinkage is, and if simply changing to a flexible epoxy will solve the issues or if there are other process changes that need to be made. Thankfully, those shouldn't be too difficult either, but no sense overcomplicating things until the root cause is identified.
  5. RideHPD

    Resin

    While my foray into lure building is pretty recent, my background is in composite engineering, so this is my bread and butter. Some random thoughts that come to mind reading others posts here: Room temperature cure resins can be quite sensitive to atmospheric temperature, generally best to work with them around 70-75F, outside of that can drastically change the progression of the cure. Microballoons are definitely a pain to handle, always wear at least an N95 mask, and better yet a respirator with particulate-rated cartridges or filter material. I haven't heard of issues for them introducing moisture into the system but a trick you can try is placing the container in a secondary airtight/vacuum container (like previously mentioned, great idea) with a desiccant like those made for gun safes. If you want to match densities to wood it's a pretty simple formula to follow, I could write it if helpful. Aways follow the mix ratios to a T with epoxies and polyurethanes. Whereas polyesters cure by condensation reactions initiated by the "catalyst", usually MEKP (which is actually an initiator and not an actual catalyst) and can be sped up or slowed down by changes in the MEKP amount added, epoxies and polyurethanes cure by addition reactions, and any additional part A or B will remain unreacted regardless of how long the cure is allowed to progress. This generally leads to a softer material that is less temperature and moisture-stable. I have seen some specify that off-ratio mixing of their chemistry results in brittleness which is odd, but they know their material. Polyurethanes also often offgas during during, so they need positive pressure to crush any bubbles from ever forming, hence pressure pots. Venting is crucial to infiltrating small female features in molds, but lightly heating the resin to decrease viscosity usually helps a lot as well. That does affect the cure process so it can take a bit of tuning to dial in without the really nice software that can model all of these things in real time. If I can answer any questions I'd be happy to do so.
  6. Oh no, I ran a thermocouple on it during one of the cure cycles and it didn't get anywhere near the Tg. If your thickest section is under an inch thick it probably shouldn't change much. I have the same, my glass bed actually has a pretty ridiculous amount of concavity to it, though these issues are extrusion related, I'm not sure what the exact root cause is but I have some suspects to interrogate.
  7. Sure thing, always happy to share that kind of info. On top of what I mentioned previously: 0.2mm initial and 0.08mm layer height elsewhere to maintain the 0.04mm incrementation for my machine, nothing significant for infill, 10-20% cubic so that all the reinforcement is orthogonal to the build plate and will provide structural support during clamping. I haven't played with line multipliers, etc. yet. That said I'm still tuning it, I randomly started experiencing some first layer issues out of nowhere last go around that I need to identify and solve, as well as a slight bowing of the top surface, so there's some residual thermal stresses to iron out, either through infill design or print settings. If there are any other specific settings you're curious of LMK and I'll pull them off the slicer. As it is I'm looking at about 21 hrs per half based on my speed settings, hence the investment into an SLA machine, but overall it's worth the time to not have to perform any post-processing on the mold, and I can do other things in the mean time. Pins came straight from McMaster: https://www.mcmaster.com/dowel-pins/dowel-pins-7/
  8. It's actually way easier to demold than with an aluminum mold. That's one of the beauties of using FDM thermoplastics for this, as they have such low surface energy that you might not even need a release agent. I've had Al molds for composites parts with 30deg of draft that were still a nightmare to demold; interface chemistry can play a really big role in demold/lift off success in your tooling design. I actually had to reduce my nominal mold release schedule for the PLA molds as it was creating defects, which you can see in that picture of the red and clear lure as the curved, recessed band around the glare coming off the lure. That's a pretty common indicator of excessive release agent across all mold materials, though I haven't really had the time to determine what exactly causes the defect to manifest that way, either from coalescing release agent, a void that's stabilized and flattened by the agent, or just really highly phobic surface to the resin that won't let it wet there. Alas, it's easy to solve, just use less release, so no need for a root cause analysis, there are way too many other things to do first. Wipe the mold with IPA, allow to dry ~15-30 min, lightly mist (if using PU) Stoner E236 or comparable with one pass, allow to sit for ~15 min and wipe off excess with clean, dry lint-free paper towel. Demold should be a piece of cake and surface quality should be top notch if you know what you're doing.
  9. I figured pictures would make this way simpler. Apologies for the trash quality, I didn't have pics on-hand so I had to rip them from messages I sent to a buddy early on in the process development. This was the first iteration with the wire locating post included as positive mold features, and also the very first casting I did out of a printed mold. Next iteration switched to dowel pins for better wire retention, as well as the printed wire jig to complement that design. And a few of the directly printed test lures themselves. Simple bores are recesses for 2mm dowel pins, larger recesses are to indicate and retain weights. If there's anything specifically you'd like to see I can take those this weekend.
  10. If you have printing experience then you're halfway there, as you know there are plenty of tweaks to be made to tune FDM printers for print resolution/dimensional accuracy. I've been doing this a lot with an Ender 3 and will probably be buying an SLA or DLP machine here soon to up throughput. I've gone about this a couple of ways with success making single body lures, but weights/etc are all fully located in the process. First way is by printing each half independently, with recesses to capture the weights and 2-3 2mm steel dowel pins to locate each half. Then you just put the weights in place and press the two halves together, using the pins to hold them together. Getting a really perfect first (couple) layer(s) is pretty key to avoid warping, elephant's foot, etc. and you'll likely need to play with your model to get proper fit of weights/pins depending on your slicer and how your printer works. 0.2mm nozzle all-day. Takes forever, but the surface accuracy is worth it. To get a quick and dirty proto I'll print the eyelets too, which requires some finesse to attach split rings/etc. too, and the action may change from the new degree of freedom you get from the flexibility of the eye, but it has some utility. You can also go a step further and include recesses for a wire harness, which leads to the next method... Directly 3D printing the mold. There's some literature saying that you need a filament with a high heat deflection like PETG or TPU to handle the exotherm, but if the volume is small you can use PLA no-problem. I've been doing that after I gave up trying to get my machine to print PETG out of a 0.2 nozzle and just sent it. Wire harness is held in place by dowel pins of diameter nominal to the desired eyelet size, with recesses to accommodate the wire eyelets themselves. This has the added benefit of simultaneously creating locating features for mold lockup, but can make for a more difficult demolding experience as you will need to design one half to a sliding/clearance tolerance to be able to demold, and resin will work its way in there. Though as long as a good mold release is used, I haven't had an issue with this, much less than with CNC Al molds of equivalent tolerance. Just don't have the pin enter into the clearance toleranced hole more than one diameter of the pin in question and you should be good. Additionally you can place locating pins external to the mold cavity to avoid this and use positive mold features to locate and retain the wire harness, but they won't retain the harness as well which may be frustrating, especially if the harness isn't perfect. That said, you can do pretty good for yourself by printing wire bending jigs to complement, so with some technique it's not as big of an issue. I'd still probably recommend using pins to hold the harness, all-said. You can do a lot of different things to incorporate weights into the wire harness, crimping lead, feeding bullet weights over the wire and retaining with crimps/thread/thin wire/etc. There are a lot of options, just takes some creativity, but can be done very easily if planned well. Ultimate production is CNC laser cut stainless plate with locating and retention features, which I do as well but is outside of most budgets ($500+ starting if you have friends) unless you're planning to go the full production route. For molds, 1mm+ top/bottom/wall thickness, equal top and bottom thickness to avoid any part warp, I'm sure I'm forgetting some of the specific print settings that help.
  11. Look around at whatever hobby, craft, and/or woodworking stores you have locally, I'd bet you can find a small molding kit for well under $50 or someone who can help you put something together under budget. For instance: https://www.michaels.com/alumilite-amazing-mold-maker-16oz./10673361.html In the spirit of making every one of your dollars go as far as they can, as others have said, trying to cut a few corners repurposing material to make molds, etc. will end up costing more money overall. Not uncommon to burn through more material trying to make the non-purpose built product work when the initially-more expensive product will likely work with much less trial and effort. For you, finding a small enough volume to get started cheaply is the best route to go IMO. And make sure to read all instructions and safety data sheets before buying anything, particularly what personal protective gear is recommended. Some of these materials can be hazardous to you, and while you'll be dealing with such small quantities working outdoors is probably good enough to be OK, some of them can have acute skin and respiratory effects if something goes wrong. Particularly so on bodies that are still developing, much the same as how alcohol can. Always good to protect your body when you're young.
×
×
  • Create New...
Top