MODERN CASTING Magazine Webinar, "Combining Cores With 3-D Printing--Pay Now or Pay Later."
"Combining four individual cores into a single piece can have a huge impact on downstream costs. It is often difficult to determine what “not having” 3-D printing can cost. In this webinar, a metalcaster will share a real world example of how progressive thinking can save money downstream.
Since 1936, Prospect Foundry has been a market leading supplier of complex cored gray and ductile iron castings in low to medium volumes. It combines modern casting technologies, engineering expertise and a focus on craftsmanship to produce high quality, complex pressure tight, porosity free castings. Prospect Foundry castings have been used by major OEM customers in the construction, mining, agricultural equipment and industrial markets."
Here's a transcript of the presentation that includes comments on foundry core coating application, pastes, patches and washes. Enjoy!
Okay, good morning and welcome to today's webinar, Combining Course with 3D Printing, brought to you by ExOne. I am Shannon Wetzel, managing editor of Modern Casting, and I am excited for this webinar, which will offer some real-world insight into 3D Printing from Prospect Foundry. Let me introduce you to our speakers. We have Terry Senish, director of technology applications at ExOne. Along with Greg Colosimo, metallurgical engineer for Prospect Foundry, and Bob Gruidl, quality manager at Prospect Foundry. Greg and Bob have a combined 80 years of experience in the metal casting industry. Thank you again for attending today's webinar and now we'll pass it off to Terry to get us started.
Terry: Thank you very much, Shannon. I appreciate it. Again, thank you, Greg and Bob, for certainly allowing us to present your story here of your experience with 3D printed molds and cores. But before we start, we'd certainly like to mention, often at ExOne, we get asked, what does it cost to 3D print a mold or a core? And sometimes a customer will show us a picture of a part. Generally we'll respond in some sort of cost per cubic inch and they may do a quick calculation in their mind and say, "Okay, well, you know what? It sounds like you're more expensive."
Unfortunately, we're disqualified even before they get a chance to see the actual downstream benefits of 3D printed cores. Often, when a foundry implements a more traditional approach and may be running into problems or issues, it can be viewed as, "Well, you know what, those are just typical problems and issues that we see and the scrap rate is a little higher, so let's just continue and proceed down the road," when they really didn't even get a chance to really see the downstream benefits of 3D printing. Really, you know when you're saying how much does a core cost? You can also ask yourself, what does it cost not to have 3D printed cores or molds implemented in your process?
To set the stage here, what I'd like to do is just talk about a real-world scenario with our customer, Prospect Foundry. But basically, if we roll back probably a good three years or so, there was a customer that was looking to produce a gray iron casting of a hydraulic manifold casting that they're looking to produce. They have some pretty complex internal features in it, so certainly the customer contracted the foundry to produce the casting for them and unfortunately, after repeated failure and limited or unacceptable castings, certainly the customer is obviously a bit concerned now and becomes late with their customers. They have to make a decision here to get their parts in time, so unfortunately - or fortunately, however you want to look at it - from Prospect Foundry's eyes or the customer's ... they pulled a job from the customer and it was from the foundry A and give it to Foundry B. Foundry B happens to be Prospect Foundry.
At this point with Greg and Bob, certainly join in at any time, but at this point with Number 5 here on the list here, you guys are under a bit of pressure right now, right? So you have time constraints that are looking to be met and certainly, in the end of this whole thing, you ultimately end up reducing scrap rate using more of a traditional approach where the scrap rate was around 70%, down to about 5%, as you guys get rolling with 3D printing. That seems like a pretty significant reduction in scrap from what we can see here.
So tell you what, Bob and Greg, I'll let you go ahead and talk about the core here and what we're dealing with and what type of problems and issues and concerns you might have had initially, maybe when you were looking at this for quoting. I'll let you guys take it from here.
Bob: Good morning. This is Bob Gruidl from Prospect. Originally, we receive this design, this part number, and we started making castings. We had a lot of problems with trapped gas and scrap on this job, as Terry mentioned. So we tried numerous tests, trying different things with venting and gating systems, pouring temperatures, pouring times, etc., with limited success. On top of that, just as you look at these cores, there are four separate cores that are made and then assembled together, so there's quite a bit of time in producing the cores, assembling the cores, drilling the cores, dipping the cores, muddying the seams for assembling the four pieces together, very time consuming, very high in cost. And then just to turn around and scrap 50 to 70% of those castings every time was just driving the part further and further ... it made us realize why the previous foundry was no longer trying to produce the part.
Terry: You guys were aware that you'd probably be running into some alignment issues and things like that, probably when you initially started looking at this core, I would assume, with your experience, if we take a look a little deeper here.
Bob: Right. With assembling four separate pieces of the core together, you have some variation dimensionally. So Greg started doing some research and came upon your process here.
Greg: Actually, first of all ... This is Greg Colosimo. We tried several different traditional type methods. We used ceramic-type sands on the shell machine. We tried zircon sands on the shell machine. We tried different coatings, alcohol-based and water-based and some types of aluminum wash. We tried zircon-based wash. Just about everything in the book. Some of the problem that we had was that the paste in the joints would swell and cause misalignment. So even if we did get the part right, it wasn't always lined up perfectly and the customer had to reject them because they couldn't drill the holes properly.
Terry: We looked at the core and certainly, it's a quality pattern making job that you guys had done here, but like you said, as you're having core paste and wash, you guys are running into some issues down the road.
Greg: Right, keep in mind these guys were doing 800 pieces at a time, so you had multiple guys on different switches, third shift, second shift, third shift, all different people. And different guys have different work habits. They have different techniques, definitely. Where one of them might make a perfect core, the next one might make a just about perfect core. The measurement of how much they were off is almost indiscernible by the eye. You'd have to measure it with a micrometers to see.
Bob: There was some variation there and that's normal. No matter how well the core box and pattern are made, you're going to have some variation. With the 3D printed cores, we're taking that variation out. We've also taken out the need to produce four cores and assemble four cores and handle four cores. Getting one core that is virtually identical time and after time, core after core, was fantastic. We also passed this information to our customer. They were very pleased with our looking into this process from a repeatability standpoint, from a dimensional repeatability standpoint. It wasn't something that we were trying to hide from our customer. We wanted to be very open about it, explain to them the benefits of it and they were very pleased with that also.
Greg: The big benefit that I saw, from my point of view, was that the shell sand has a 2.7 to 3.4% resin whereas the printed core, it's just over 1% so it was a lot less gas being generated.
Terry: Again, looking at this more traditional core here that's been pieced together, you had mentioned about all the different, I guess, idiosyncrasies that you can have in here and certainly that the glue has to be dry as well, too, right? From a paste perspective.
Terry: You guys are almost making it sound like there's more than one way to screw up a casting, right?
Terry: So now, once you start with your traditional approach here, and you set these cores into the mold. You guys are probably under a little bit more pressure right now to get this done for your customer because he's still late, isn't he?
Bob: Correct. And that was the other thing. The timing to go through all of the ... it was a hot item with the customer because of the lack of success from the previous foundry. We were kind of under the gun from day one and then finding all the issues and all the problems and just having enough time to produce and assemble, as I mentioned, all the cores, just put a tremendous amount of pressure on us. You'll see in some future slides that you have here, the problem with the castings, all the trapped gas that was in there. Greg related to that ... yeah, there you go. That's some of our earlier castings that we made with our cores. All the gas that was generated ... because, you know, you're pouring iron 25-2600 degrees into a sand mold and surrounding that small little core was all this molten metal. You're creating gas and that gas has to go somewhere. With ExOne's process, in discussions with Greg, between the two companies we came up with the best solution and the best type of core sand to really pinpoint and eliminate our main problem that we had.
Terry: Again, Foundry A, who obviously went about this in a more traditional way, using traditional patterns and again, when it got turned over to you guys, Foundry B or Prospect Foundry, again, your initial knee-jerk reaction was to go about it in a traditional way, to say, "Hey, let's create our patterns and move forward." As far as even considering 3D printing early on, it was really not in the front of your mind to pull that wild card out, wasn't it?
Bob: That's correct.
Terry: It wasn't until you guys had to go outside and find some additional solutions here, and a little bit of a panic, I would assume.
Bob: Not quite a panic, but you're on the right path.
Greg: When I went to the ASF regional 2015, one of my key jobs there was to find some kind of way to fix this problem. I talked with a lot of vendors about core coatings, about non-traditional types of replacements for sand to create the shell sand to make a core out of. These were very expensive, very expensive solutions. I couldn't see that it was going to create something that was very much different from what we already had. Then I talked with the gentleman at the booth at ExOne and when we finally got around to the point of finding out how much it would cost to make these cores and compared to our limited success, just all the work that we put into it, we found that it was actually cheaper to get the cores printed. It cost us less.
Bob: Right. That's not even including all the hidden costs and all the other potential lost confidence from your customer, or potential of work because you can solve their problem. I could go on and on about that.
Terry: Right. Basically, if I recall correctly, at the trade show, which again was just a mini regional show up in Wisconsin, when you had stopped by, I believe we had a sample core on the desk. You said, "Boy, I wish you could print a thousand of these for us," right?
Greg: Exactly. You actually said, "Yeah, we can do that for you."
Terry: So, what, you made a phone call right there at the foundry to your shop to see if you could combine some cores, right?
Greg: Our foundry engineer had put together the model for the cores. Basically, what he did was he got the model that they used for the show and he combined those. He got the core prints off of those and then he stripped away the metal part and came up with the inside passageways and he combined them together and made the actual core model.
Bob: The model that you currently see on the computer screen.
Greg: That's that model right there. I had that sent to the foundry to ExOne and they said, "Yeah, we can make this for you." They started on the samples. I think we sampled four different types of resins and sand, and we finally arrived at the best solution with the help of the team at ExOne. From that point on, we actually thought we were going to lose money in 2015 because of this job but we actually ended up just about breaking even. The next two years after, we've done very well with these parts, very well.
Bob: The benefit for getting the different cores ... it's the same core but the different process, different types of sand or the amount of resin. The reason that's important is because different part numbers, different foundries and different part numbers in those foundries, you might be fighting a different animal. Are you fighting an animal where the assembly time is killing you? Or are you fighting the enemy that the trapped gas is killing you? Or some other thing? Or the cores are breaking or etc. You could go on. In this case, it was not only the assembly time but the LOI and the trapped gas and the mold. We really got a double punch, a good double punch of saving time on assembly and improving that LOI so that we reduce the amount of gas that is created in the mold in the first place, which is a win-win on both situations. One type of core across the board ... it's good to test different things, maybe is the point of that.
Terry: Right. So when we run this course over at ExOne ... again, we've printed thousands of cores for you guys at Prospect for quite some time. A couple things. Again, we've kind of initially solidified the furan and silica option, which worked out well and you guys applied an alcohol before wash to that, which appeared to work best of the different sands that we have tested and binders. But I think ultimately what people should understand, too, from a ... ExOne adoption centers, we certainly do our best to maintain high quality cores as well too and those that have machines, certainly try to do the same obviously. One big thing is the 50% humidity that we try to lock in our shop here, plus or minus a few points here. But nonetheless, that allows us to use a consistent amount of binder and activator throughout the year and have lower LOI, higher strength. As you had mentioned, Greg, your LOI was what, initially, with the traditional cores?
Greg: We had over 3% LOI with the traditional cores. You know, Terry, we don't have problems at Prospect making cores. We have 14 shell machines. We have an ABC set up and we also have furan for larger cores. We're very proud of our ability to make difficult cores. So it wasn't like we had one hard one that we ran to look for a solution. This one just wouldn't stick.
Greg: For people that have shell machines, they know that sometimes there's cores that have what they call false fill. It's when you get a layer of sand on the outside and it's just a fine layer. And it collapses when they pour the metal in and then you get a scrap casting. Well, you know, that was another thing that I was thinking of that 3D printing would cure that, because we have a couple jobs that we might be looking at where we have this problem of false fills. It takes extra time because they all have to be checked and patched and then recoated before they go into the mold. That's the other thing 3D printing could easily fix.
Terry: Right. Again, there's a lot of things, a lot of customers ... if there is even a slightly higher scrap rate of anything ... again, there's a certain amount that's considered normal to the industry. You know certainly some experimentation with 3D printing, you know, may find some downstream benefits that normally you would not normally see or uncover. You know, one of the biggest things is shipping is generally not our friend when it comes to 3D printed molds and cores. We've gotten very good at it over time and I know you'd mentioned that our initial packing strategy was not as efficient as what we do today. Certainly we've gotten good at shipping cores over the years. If cores are too lacy, what we can do is ship them in a box with loose sand in there to help support the core during the shipment process. It sounds like, though, with everything that you guys have received, though, you guys have been pretty happy from a shipping standpoint.
Bob: Absolutely. This particular core has small passages, as small in diameter as the passages are, is about as fragile as you can get. You get big heavy core prints on the end and then you have all these quarter-inch or actually slightly below a quarter-inch running through this thing. They're very susceptible to cracking and breaking. The packaging that you send them in is fantastic. We very seldom see a cracked or broken core. We also run our own LOI test here at Prospect so if we don't find a cracked or broken one in the box, we end up breaking one ourselves just so we can our internal tests.
Greg: That's right. That's one thing that's been through day one, since the first shipment. There might have been a lot of foam in there but we still didn't see any broken cores. That was one of our initial worries, that when we talked about ordering these, the guys in the office are saying, "Well, how are they going to ship these cores without breaking them?" Because we bought larger cores from other vendors and when they arrived, the rubbed against one another or in some way, broken. So, that was maybe a worry when we first purchased cores from ExOne, but we haven't had a single time yet when we've had broken cores in the boxes.
Bob: It was a worry that was not needed.
Terry: Great. That's good to hear. I guess you're right. Whether it's the traditional blown core or a 3D printed core, you know, they certainly got to ship them and get them there in one piece. Yeah. This is kind of showing your foundry there and the finished part that you produced there. It sounds like your customer has been pretty happy here, when it comes to the casting, the shipment delivery.
Bob: It's a success story, no doubt. It solved our problem and it made our customer happy. It turned a losing job into a winner.
Greg: I think we've made over 3,000 castings this year so far to September, so we'll see. We might hit 4,000.
Terry: Great. That is excellent. The question, what does it cost not to have 3D printing ... certainly you can try to assign a dollar value to that, but what do you guys suppose Foundry A initially that got the job pulled from them ... This is thinking ... I know you'd mentioned you don't want to be that guy, right, when it comes to a foundry that could not deliver. Just curious how you ...
Bob: If you look at what does it cost ... Let's just start from the beginning. First you have to rig the tooling to run in your ship and that costs a few thousands dollars to get the gating system and patterns ready to run. Then if you run a test, you're breaking the production on whatever molding line you're on and you run your samples. It probably costs anywhere from 1,500 to $2,500, depending on what the machine is, just to run that test process samples and to have people cut up those samples, inspect them, etc. etc. Say, you run 5, 6, 7 different tests and you're still struggling with this job, just on an average figure, that's a $2,000 bill every time you're doing that. There's a lot of lost revenue. You could be producing something that you could sell to a customer at that same time you're running these tests that are failing. There's a lot of money there. Then, just the frustration of it all and dealing with your customer, as we mentioned, and you know, trying to make promises based on when you think you can get this job up and running. It leads to a lot of lost revenue that you can put your hands on.
What does it cost not to have it? That's a great unknown. You really can't put your hand on it because you don't know how your customer is going to react. They may be happy that you're trying to help them with a known problem job that's already failed somewhere else, but at the same time, you know, if you solve the problem, they're like, "Hey, great. These guys figured it out. They know what they're doing there. We need to get more work into their foundry." That's that terrific thing that you don't know, but if you don't get it, you'll never know. If you do get it, you can measure it because you just measure your increase sales dollars with that customer and there's your success. What I can tell you on this particular part number is that the first year, we were looking at around $46,000 in sales. Well, it's gone up since then. This year alone, we're at $138,000 in sales on this one part number. You know, that's just one part number. The fact that it's profitable for us and it's economical for us to do it, getting the cores produced by ExOne, transporting them to us and getting them to the molding line is economical. It works. It's repeatable and our customer couldn't be happier with the product that we're shipping to them.
Greg: Another foundry could be saying, "Hey, we could have thought of that," but they didn't think of it.
Terry: Right. Again, I think, is 3D printing great for every application, from a cost perspective? No, but there are times when it can be. But nonetheless, when you guys are ... foundries in general, when they're running into some problems, certainly it's often worth testing and trying 3D printed cores because some of the downstream benefits, from what we see in other industries ... for instance, the pump industry. That with one piece cores, what they're able to do is reduce the spacing and tolerance stack up and eliminate downstream balancing issues and have better-performing pumps that arrive at foundries. Again, a lot of these things, unless you experience it, it's something that is often overlooked, I think, moving forward. But yeah, again, Greg and Bob, I again appreciate you sharing your success with us and certainly, I think what we can do is open up to questions for anyone that would have questions for myself or Greg or Bob at Prospect Foundry.
Hill and Griffith Customer Service
The Hill and Griffith Company's green sand metalcasting foundry supplies help achieve the EPA's M.A.C.T. standards and reduce Benzene emissions. Our variety of environmentally sound release agents, coatings, partings, lubricants, core oils and specialty products will help you meet your metal casting's needs. We're known for our hands on approach. Let us visit your plant and recommend products that suit your needs.
We are pleased to provide samples in quantities large enough to allow you to "try before you buy."
Contact Us »
Technical Services & Support
On-site casting defect investigations, product testing, machine start-ups and much more. Also, lab facilities are available to provide testing upon request.
Contact Us »