In July 2021, Brian found a website called forums.thehomefoundry.org, run by a man named Kelly Coffield. Kelly also has a YouTube channel. What Brian learned from Kelly's videos led him to make two changes to the grail building process.
The first change is to adopt Kelly's particular approach to the lost foam method of casting.This means using the same tools that he does. Petrobond sand will be replaced with (if I'm not mistaken) regular sand. Instead of building a box to place the grail blank, sand, and molten metal in, a steel drum will be used as a flask. This steel drum will be placed on top of a dolly (i.e. a metal star or cross-shape with a wheel at the end of each leg/prong) which has a vibrator and a vacuum connected, both of which are air-powered. Note that the vibrator will not be used for the grail, but it will be used for other projects (therefore, the dolly might as well be built with the vibrator attached). Additionally, Kelly recommends buying an industrial-grade slurry in bulk for the best results.
The second change is in regards to the adhesive used to connect the foam plates which comprise the grail blank. Brian would like to switch from the yellow expanding foam used in his grail building videos to a glue recommended by Kelly (specifically for foam), Aleene's Fast Grab Tacky Glue.
long as lost foam remains more viable than lost wax (basically due to
the large upfront financial cost of the latter), Brian wants to
implement Kelly's methods for other projects as well. Another article
will posted on further details.
The following information is Brian's e-mails to me on these matters.
Change #1: Lost Foam Method Tools
E-mail from Brian #227, July 7 2021 9:47PM
Well now... I've got some interesting news myself. I've stumbled upon a genius. I'm damn near giddy with excitement about what I've learned over the past couple of days. It's taken all the guesswork out of the processes of the lost foam method of making a grail, and MANY other things... including the Di Pietro engine, and just about any other mechanical device I've spoken about, at least in the realm of lost foam. I've learned also that Petrobond sand is not necessary, and my forceful nature of trying to implement it into the process was a waste of money. Literally, other than the making of the grail blank (which is still the same process), EVERYTHING has changed. The man's name is Kelly Coffield. I was watching "machinist porn," and trying to find a variance of a superior shop grinder while daydreaming about the tools in a shop dedicated to our projects, when his videos popped up on my suggestion list. The grinder videos are awesome in and of themselves. Watching those types of videos brings me a calming inside that I don't seem to be able to get any other way, but what happened after watching them got me downright excited. The grinder was designed and made by a guy named Phil Vanderlay. He has some great stuff on his channel, and he's willing to share all of the plans. Plus he has that machinist wit that brings a lot of humor to the equation, at least from my... well I guess now OUR perspective. I'll link his series of videos on the grinder build first. You'll enjoy them, especially if you enjoyed the my mechanics stuff I told you about months ago.
He also did an improvements video where he made a dust collection system to shield the sparks. That's the video where I saw the suggestion to this video...
I was extremely impressed with how the fins on the heat exchanger section of the piston housing came out without needing refinement. I've seen this done before, but it always seems to be in a high production, factory type setting. Not exactly useful to a hobbyist, which is more the route I'd like everything to go, but this guy, this genius (and I don't say that often), showed me a method I had never even considered. The tool talk was interesting, and I've seen many different ideas for how to compensate for complex issues in making variable shapes, but at the end of that, where he shows his molding apparatus... I was shocked. Absolutely fucking genius! At 7:33 he goes into an explanation about his self conceptualized dolly, complete with an air powered vibrator and vacuum port. At first I was confused. I've thought about using a vacuum for molding on several occasions, but I had never even considered drawing it out from the bottom of the flask. My mouth was agape as I sat here last night, dumbfounded by the absolute fucking genius I was witnessing. I can't stress enough how significant this is. I knew it as soon as he started explaining it. I literally put my phone down and went for a walk in the middle of the night to think it through. I even spoke to my dad about how excited I was. I told him "this has taken all the guesswork out of pouring the grail." And it has. All the variables that brought Shane and I angst, all the gating strategies, the lazy Susan idea of using centrifugal force, the hours and days we spent debating how to make sure that the metal would fill out the void left by the disappearing foam blank... were all unnecessary. Even the Petrobond sand is unnecessary using this method. The sand used for this method is also 100% recyclable, which the petrobond sand is most definitely not. The amount of efficiency this man just injected into the entire (and not just for a holy grail) process is truly remarkable. Everything being done with lost foam... worldwide... should be done using this man's conception. While going for my walk trying to calm down from the excitement, this was what I was thinking about.
My mind was racing with ideas, but I also wanted to study everything I could about this guy, so I came back to watch more of his stuff. Unsurprisingly, he has done a few experiments to test the boundaries of this capability, and that's when I saw this series of videos...
This was the series that showed the vacuum he made, and the dolly in action. It works just as I thought it would. Interesting how he was aiming for 1/16" thickness, and the grail blank I made was 3/16" thick. It shows that with even just aluminum, which has a lower melt temperature than bell metal bronze, that thickness was still achievable. The gating system that had to accompany the setup for success is not really a concern because I already knew that might be an issue with the grail, but seeing this process in this way eliminated a significant amount of doubt and questions that arose from that doubt. I haven't been this impressed with someone's ideas since Shane. Watching Shane overcome the immense pressure from all sides to conform to a particular paradigm, and just follow the herd mentality, especially considering where he was raised was a beautiful thing. This guy gave me that kind of hope.
There were several videos of his I watched last night, and fell asleep reading his forum posts talking about improving the already fantastic design. I'm going to spend the next several days studying everything I can about what he has to say. This process can be used for so many applications, I'm kind of discombobulated by my thoughts right now. One of the videos I watched last night had him discussing what he calls a pin router. It's an interesting device that he uses for more complex shapes, and he's done an amazing job utilizing jig setups for large scale production. That video is here...
While I do see a great many uses for this type of device, I see better application (especially for our projects) in the usage of the devices I told you about a while back: the copy carver, and the pantograph router (pantorouter). Using Kelly's pin router is a great way to make 1:1 scale objects, but the pantograph router allows for scaling up, or down. Any ratio is achievable with a properly balanced pantograph router. I see immense potential now for the Di Pietro engine to be scaled up using this method, and the best part is that with the precision implements that this dolly system provides, very little if any refinement after casting is required. Additionally, the ability to machine foam for casting (or wax) from a smaller scale, then casting it in metal saves the tooling itself from damage. The end mills will last a lifetime. Extremely efficient use of expendables. Imagine building a Di Pietro engine that's less than 6 inches in diameter. Then using a pantograph router to upscale the exact motor you just made 10x the size of the 6" motor... and doing that exact upscaled replica out of foam, then casting it. No more need for CNC at all. No need for large stock for machining. Literally everything just got immensely more efficient. The entire process. Once the routers are set up for air pressure or steam, no more need for electricity. A fully functional CNC machine shop, where the human plays the part of the "C" in CNC, completely off grid and impermeable to any type of external influence. My dreams... and they just got a lot more efficient.
As you can tell, I'm very excited about the possibility. Also am somewhat disgusted. In a world where a used car salesmen schlep like Elon Musk is worth billions, and a man like Kelly Coffield has 1000 subscribers on YouTube is the norm, disgust is inevitable. I'm going to get back to reading and watching everything I can, and I'll try to relay to you everything. This is the forum posts I passed out to while reading this morning...
There's a lot of technical stuff in there about the new device he made. I'd love to be able to meet this guy, and if there's a chance that we're able to create a makerspace type of university-like environment some day, he's at the top of my list for being an instructor.
Anyways, I'll catch up with you later after more research. Hopefully this helps make you feel a little better.
#229, July 11 2021 1:07AM
Just to be clear on one other aspect of this, the vibrators wouldn't really be necessary for the grail. It's a pretty straightforward design, and with the invert-able idea (which was always going to be the chosen method), the filling of the internal area would not really be a concern. The vibration would make casting engines and other applications with voids and ports much more efficient, though. Anyways, hope that helps.
Change #2: Glue
#238, July 20 2021 11:07PM
Okay... I'm going to officially make a change to the holy grail blank machining process. As I've explained several times to dozens of people, all while trying to do as much research on this subject as humanly possible, while also spending hundreds of dollars trying to locate the best method, Kelly Coffield once again comes through. Seriously, where was this guy two years ago? Anyways, I'm going to say once again that his superior knowledge on this subject has given me the confidence to say that his system is better than mine without having to confirm the material myself. He recently made a video that I just watched for the first time right now where he gave a solid breakdown of his lost foam "blank" materials, and explained the pros and cons of every material. It's a long video, but the information about what glue he prefers is around the 9:51 mark...
It seems as though he calls his "blanks" patterns, by the way. Personally, I've tried dozens of different glues while trying to find the best one, that doesn't destroy or warp the foam. This was arguably the most difficult aspect of trying to make the lost foam method work for this kind of application. After watching his videos on machining his patterns/blanks, I'm confident that this glue will work for making a holy grail blank. He does some pretty intricate patterns with fairly thin walls where joints are put under considerable stress while machining. Also, a major benefit to using this type of glue, as opposed to the expanding foam from my videos, is this will dramatically cut down on wax. I would still recommend using wax to provide the best overall surface finish, but filling the voids left by the expanding foam would be a nonissue.
I would like to say that I did try wood glue, and it did not hold up well to machining. That could have been the fact that it wouldn't penetrate the foam surface, or it just wasn't strong enough in those thinner machining applications, but I did try it. I was not impressed. Flying blind with this type of glue, other than Kelly's superior knowledge as a guide, I would still do everything the same exact way. In other words, sanding the smooth factory surface finishes before gluing the sections together is still highly recommended. As well as putting the plates together and holding them in tension with a tail stock on the lathe while the glue cures. Figuring out how long that cure time will be will depend on the environment. From the sounds of it, based on Kelly's information, it shouldn't take any longer than an hour or two, but that might vary. Temperature, humidity, air circulation, etc will play a factor in the cure time, but just based on what I've seen, and what Kelly has explained, switching to this glue will not increase the total duration of time it will take to complete a finished holy grail blank.
This is the first official change I'm making in the grail blank making process. It'd be nice if I had the ability to recreate the blank and videos, but as we both know I am not in the position to do that. I am considering making an updated description to include this information, though. I'm strongly suggesting that you take this advice, and do it Kelly's way... if/when you make your own grail blank. I have no problem admitting that Kelly is more knowledgeable than myself in this medium, and I apologize for not having figured this out sooner. I tried, really I did. I thought that I had found the best possible glue available for that application, but apparently I was wrong. I'm sorry. Anyways, it's a good video with several areas covered that we've discussed in previous emails. You'll like it. Talk to you later.
#239, July 21 2021 12:01AM
I've been reading reviews about this stuff, and haven't really seen anything negative. Apparently it's at some Walmart's and Michael's but I don't recall ever seeing it. Might be in the crafts section. Here's the glue Kelly mentioned. I'm bummed out that I don't have the means to try it. It really does seem promising.
I did read one review that said it dries somewhat flexible. That shouldn't be an issue if it's put on thin, but it was the only concern I've seen. That might become an issue while machining the opening. There's a chance that the flex might fuck up the concentricity and/or make chatter harder to deal with. Just trying to cover my bases in case you encounter problems. Still, though, the expanded foam "glue" also adds issues, so it might be a give and take kind of operation. There will almost definitely be pros and cons to both methods, but after studying this guy and seeing his results, I trust his judgment enough that I'm confident to say his way is better... for the lost foam method at least.
Why This Glue?
#240, July 21 2021 8:28PM
I tried Gorilla glue. The offgassing from the curing process eats foam. It's not supposed to, but when used over a large area where the outside of the curing area dries before the internal areas, there's nowhere left to displace the curing vapors. That's the basic problem I had with every glue. The various forms of "super glues," like Gorilla all had that problem. After several failures in trying to find one that wouldn't offgas and eat foam in the process, I bought this stuff...
3M 78 polystyrene adhesive, which is specifically formulated to NOT eat foam, also ate through the foam while curing. That stuff is not easy to find, either. I had to order it online and waited for a week just twiddling my thumbs before it arrived. I was excited when it got there, but extremely let down and disappointed when I saw the results. Plus it had the same problem as most of the super glue products I tried, and Kelly hinted at that in his video. None of them fully cured to a solid state where machining was even possible without gumming up and wrecking the tools. Without a dry, clean tool, machining anything prone to chatter is a nightmare. Not only does it wreak havoc on tooling, but makes sanding all but impossible. I even tried just sanding with no tools. Nightmare fuel. There were several times where the sandpaper stuck to the blank and was ripped out of my hand causing huge gouges and throwing concentricity way off rendering the entire blank useless at that point.
That picture you sent me looks like an aerosol can. That's no good. There's solvent added to the glue to dispense through the tips/nozzle... and guess what? That shit eats foam. I'm sorry. Maybe I should have been more particular about that excited rant I went on last night. NO SOLVENT, NO AEROSOL, NOTHING HEAT ACTIVATED. Sanding heat activated stuff reactivates the glue then you run into the same problem I had with the super glue. The real issue you're dealing with on this glue issue is two part. The first problem is that on those first couple of plates, the entire surface needs to be covered and sandwiched with another plate. That makes offgassing a significant concern, and right off the bat. The second problem is making sure that the plates can move around a bit while you center the next plate. Gorilla glue was a problem for this as well. Kelly also spoke about this in that video. You don't want instant strong adhesion. You want it to be able to be slid around, spread out, and pushed into place using tension while you do that. Apparently, according to Kelly (and he showed this with the Aleene's stuff which made me very happy), that product does that very thing.
The next situation that needs to be addressed is how it cures. I can't stress enough that the glue used needs to be solid and uniform along the entire surface. Without that happening, machining is all but impossible. But moreover along this ideal is the curing time. Technically you could wait an entire day or two between plate machining sessions, but that would mean a couple of weeks just to make 1 grail blank. That's not very efficient. Possible, yes, but it would get very annoying waiting between glue curing times, and also make cleaning up the mess repetitive and loathsome. Where the expanding foam was great in this regard, was I could do 3-4 plates a day, and only had to clean up at the end of it. That dust and shard mess is significant. That shit expands outward vigorously and sticks to everything. There's static involved, and during sanding, the particulates are very small. Even if you didn't mind going through those motions every time you machined an individual plate section, that's still a long time between completed grail blanks. Imagine if there was a problem during the pour. You're two weeks out to make a new grail blank. For me, especially when Mangum was nearing the end, that was not possible. If you were planning on making several blanks to combat that problem, keep in mind that the lathe itself is the vise/clamp that's holding tension while the glue cures. That's done so that concentricity remains stable enough to machine plate by plate. If you took the backplate off of the chuck to mount another blank, you might end up with a very wonky grail blank by the end, or you'd be creating a lot of extra refinement machining every time you mounted the backplate back on the chuck. Which would ultimately create a time loss that would nullify the whole idea of saving time by making several blanks at once.
I'm aware of the problem expanding foam adds to the equation. I watched several videos and did a lot of research on this before I tried it. Ultimately, after all the other problems created by the various kinds of glue I tried, the expanded foam was worth the risk. This was the video that sold me on the idea...
It's not perfect, and Kelly did mention this problem in his video. Also, one of the only comments on my videos was about this very issue. The expanded foam does have a higher melt temperature than the polystyrene. Given all the other issues, it was still the best solution to the problem. Gating strategies were discussed to compensate, mold preheating, etc, but what I ended up doing was trying to put as much pressure as possible on the plates while the expanded foam cured, and make that layer of expanded foam as small as possible. That's why those layers between plates are very small. It was a challenge trying to fit all of this information into the videos I made. Nathan and I discussed the length of them and how much could be edited out because they were so long. Even though I don't talk in them, I tried to show all of this stuff in practice in real time. Some of these things are probably just glanced over while watching the videos, but every little thing in them has a backstory, and this glue thing is by far the biggest hurdle I had to deal with. Make sure... for your own sanity's sake... if you're going to try to Aleene's stuff, MAKE ABSOLUTELY POSITIVELY SURE, you're using the EXACT stuff that Kelly suggests. Otherwise, just stay with the foam.
It's interesting, but I'm getting kind of nostalgic talking about this stuff again. For weeks on end I was talking to Shane about all these problems. I also tried talking to Robert about it, but it seemed like he just didn't care, then got that fish mouth, drooling, fluoridation stare every time I tried to bounce ideas off of him. Shane was much different. I can't tell you how many times we would be ranch handing at his farm talking about this stuff for hours getting dirty, sweaty, working in the fields, drinking the last bottles of my beer (The Lord's Brew Hefeweizen [heh, that's really what I called it]). And he would do that with me after full work days. I really do miss that guy... I don't think I've known anyone like him. So willing to work hard, and the hard work itself was seen as fun. Plus working-ish/contemplating while we worked. He is definitely rare. Anyways, this is a pretty detailed message for such a small question, but I'm not sure if I've ever really gone in depth about the "why" behind this stuff outside of the Shane discussions. Hopefully it helps you.
#242, July 23 2021 1:12PM
This glue problem is a nightmare to deal with, at least for lost foam. There's so many factors that cause problems with the foam, and lost foam is such a niche procedure for "back yard hobbyists" it's difficult at best to get accurate information. That's why I respect Kelly so much. Even before I saw that most recent video about the glue, I could tell he got his information through direct experience. These days, you never really know if people are giving you second hand information, or they have some financial vested interest in the product, or just simple pride. That's where guys like us differ. We just want the most useful and efficient process to get the job done. It's amazing how difficult that's become. Energy production, food production, even things like purity, the environment, and politics. Seems like there's more misinformation than direct firsthand knowledge. It's exhausting and so unnecessary...
Yeah, I fucked up when I took a quick glance at the bottle you bought. Upon second glance I saw that the cap was on the bottom and the top of the bottle had obvious mold seams. That's not how aerosol cans are made. I just didn't want you to waste time and effort on anything aerosol. That shit is bad news. I tried 5 or more different kinds of aerosol glues. So disappointing. Same with the super glues. I tried a few of them and they all sucked. I was getting most of my information from forums and YouTube videos where guys would build models and wilderness settings out of foam. They didn't machine their stuff, and apparently covered up all the flaws in paint and shit. Those were some frustrating days.
#231, July 12 2021 12:14AM
Oh yeah. I forgot to add that there should be no divergence from the construction of the grail blank. That the most efficient way to make one: with a wood lathe, and done in sections. The only upgrade I'd make to that process is to convert the electricity to air pressure on the lathe itself. One of the more worthwhile benefits to that conversion would be the versatility and range of the RPMs. There'd be a simple ball valve controlling the air pressure, so the RPMs would be determined by the amount of CFM allowed by the valve. You could use a fraction of 1 RPM, or several thousand depending on the capability of the engine installed. Far superior to an electric wood lathe, especially the one I had. The lowest setting was about 500 RPM. Kind of became a pain in the ass when trying to center it. Also with the length of the blank itself, without a tail stock, it limited the size. A longer blank means less stable at higher RPMs. Anyways, just throwing that out there. Making a grab blank hasn't changed at all... as of now. I thought I mentioned that very briefly, but maybe I didn't.
Take it easy and I'll talk to you later.
Because Kelly's approach to lost foam casting led Brian to make so many changes to his plans, it may be a blessing in disguise that Brian did not get a chance to try casting the grail yet. Let us hope so...