You briefly mention a "shrink ruler" when you first introduced the use of the spring calipers. for the folks who don't know, a shrink ruler is set up to include the expected shrinkage when a given casting cools. For example, a shrink ruler for cast iron will be engraved to read the final dimension of a cooled cast iron part - so the pattern needs to be a few percent over-size. Other shrink rulers are calibrated for aluminum and bronze - probably a few others are made as well. This way, the pattern maker isn't always trying to calculate the over size dimensions while turning and carving - very critical in the days before electronic calculators!
Thanks for the education, can't wait for the next installment. That pattern is so damn pretty I'd be inclined to use it as the pulley instead of having one cast, but I understand staying true to the original materials thing. The machining stage will make an interesting video too.
That was soooo interesting. Is there any chance we can see a video of the mold making and casting process? I spent a month in the foundry during my apprenticeship (probably 1967 or 1968) but we didn't go in the pattern shop. So while we obviously used split patterns like this, it is long enough ago that it has become a bit sketchy. I imagine there is no shortage of casting channels on TH-cam, but that won't be the same as following this part through the process. A great video Keith.
Nicely done and presented very well. Very elaborate glue up on the pattern. Did you mention this was a museum? Are you a volunteer or a member of the staff. Either way, thanks for taking the time to make these videos.
Kary Miller Thanks - you're being silly now. Our comments are showing up on this guys TH-cam video. I don't know how this Google+ thing works. My original comment was about this fellows work.
A slightly different way to register the two halves is to first glue them together with paper between them. Then drill a holes that go through one half and into the other. Drive your wooden dowels into the holes and now the two pieces will be perfectly aligned when you remove them. This way you don't have to worry about the two pieces rocking slightly off axis when you drill the holes in each half. If you use paper between the two half of the pattern when you do the glue up you will have a strong enough bond so the halves do not split when turning but you can then insert a knife at the paper and actually split the paper. I think that tubalcain has a video showing this process.
Thanks for another great informative vid. I have two questions. Why not weld the pulley instead of replacing? Old machines are often found with repairs, so you would still remain true to the technology and save hours of work. But then again, we wouldn't see how this process works so thanks for the extra effort. I've never turned in wood, but instead of using all these segments, wasting hours in the process, why not use MDF to make up your halves. MDF is very stable to temp and humidity, extremely flat, glues well, and is easy to work on high speed shapers. Plus there is no end grain and no internal segment joint seams. Does it not work well in a wood lathe? That said, I appreciate seeing the old methods using lumber.
@@VintageMachinery Hey Keith, you did use a shrink rule. For the record, shrink rules are not calibrated for any specific metal. Rather they adjust for a specific amount of shrink. For example 1/8 in per foot. In theory a strip of iron will shrink about 1/8 in per foot. In practice that’s not always the case. Depending on the iron and the shape of the pattern, you may not see any shrinkage. The core serves two purposes. First it gives you a rough starting point for machining. Second it evens out the thickness of the casting features resulting in more even linear cooling and a reduced dependency on risers to mitigate shrink voids in heavier sections. Thirdly by making the cooling more even or linear, it greatly reduces internal stresses in the part.
If you were concerned about keeping the "drive" end stronger, and left the 1 inch core oversized to 3 inches to allow this, why not turn the pattern 180 degrees and have the thin core on the centre end, allowing a drive though the 3 inch core?
When using the joiner, the leading edge almost always ends up thinner than the trailing edge. I like to turn the piece end to end each pass to keep things even if the grain allows.
In theory this is false. In practice this is true. I noticed this abnormally on my porter 300. In terms of blow out. I am not a fan of fancy insert heads. This may be one case for them, I don’t know. But the jointer is about accuracy. To many folks forget this chasing surface finish from a jointer. A scraper plane such as a Stanley 85 would truly solve any surface issues here. A Stanley 112 would also work and it had an adjustable angle.
Gouging always worries me, with the guide being set a half inch or so away from the piece being worked plus the increasing distance away from the guide when turning, my self preservation says that the Gouging tool could catch on a Knot or a fixing screw and flip/catapult/ seesaw of the guide and take your head off, I would have thought that a second guide situated above and in front/ (closer to the operator) of the bottom guide would be 100% safer, that said I am aware that the method here has been used for years so are my concerns unfounded ???
***** Point taken Keith, experience is everything, but it seems to me that there's a heck of a lot of power/velocity which could catch an operator off guard in a split second if your concentration is distracted
I found it rather amusing that in part 1 of this project you fussed with the miter trimmer to make all 6 segments fit perfectly with no gap. Now you still went ahead and jointed the mating surfaces flat. So It was totally unnecessary to be that particular.
***** That s great, if you want to preserve the technology. Unfortunately in factories you don't work like this anymore. You use CNC or 3 D printing. In Hungary for example at Opel factory the prototype for a gear box in made with a 3 D printer. Using an older method with wood would take about 6 months. Wiht CNC is very hard to do it, 3 D is the ideal solution. It takes only 6 weeks do finish it.
Szaby Sascha 3D printing is a fad. It replaces the woodworking skill sets so an idiot can do it. Your efficiency is achieved by placing the design into solidworks or whatever your using. I machine wood all the time on metal working machines and use CNC control from time to time to do it in making infill hand planes. In terms of getting there... your either going to remove material or add material. Two different viewpoints... same end result. I am more interested in the thermodynamics of cooling metal. How does the metal flow in the mold. How does it shrink. Do I need a match plate. etc. etc. 3D printing can help you but its not as effective as you think. Your still going to need to make the final parts. Most plastic injection molds today are done in solidworks, emailed to china, and the final injection dies are fedexed back to the US. No 3D printing. Just CNC machines removing material. For what we do, this method still works quite well. Taking an old time pattern maker six months to make a pattern strikes me as strange. Are you guys training the patternmaker during that six months? It took MoPar two weeks to build a new version of the 426 HEMI engine after a significant design change. (Set of wooden chrysler patterns).
I don't know the 6 six months thing. I just read it in a technical magazine from Hungary. An engineer wrote an article about this topic, and he said that. Here in Romania there isn't any training for this profession any more. My dad learned this in the 60's , in 2005 the factory where he worked said that they don't need this department anymore.
Sadly, he is correct. The pattern world has been replaced by Solidworks which is an incredible software program. But wood was cheap and quick to machine so it was a perfect medium for patterns and prototypes. Now, we have CNC machines that can cut solid aluminium faster than slick willy. Just dont let the 3D printer fad catch you up. In some cases, its great. Usually in those cases, the reason the 3D printer can do the job is precisely the reason the foundry guys are going crazy. Patterns still have to be rammed up on a squeezer line and that means core prints, cores and draft. Some things change. Some things will never change.
That's one well equipped wood shop!
I really enjoy your videos, Keith. Thanks for taking the time to do such a professional job of sharing your skills.
This is such a great project! I go back to it again and again. Thank you Keith!
Excellent video and great information! Thanks for sharing this.
You briefly mention a "shrink ruler" when you first introduced the use of the spring calipers. for the folks who don't know, a shrink ruler is set up to include the expected shrinkage when a given casting cools.
For example, a shrink ruler for cast iron will be engraved to read the final dimension of a cooled cast iron part - so the pattern needs to be a few percent over-size.
Other shrink rulers are calibrated for aluminum and bronze - probably a few others are made as well.
This way, the pattern maker isn't always trying to calculate the over size dimensions while turning and carving - very critical in the days before electronic calculators!
WOW - Great explanation of the casting process.
Thanks for the education, can't wait for the next installment. That pattern is so damn pretty I'd be inclined to use it as the pulley instead of having one cast, but I understand staying true to the original materials thing. The machining stage will make an interesting video too.
The machining part of this pulley was done by Keith Fenner in his series "Working the curves" Part 1 was released Jan 20, 2014.
Very, very interesting !
That was soooo interesting. Is there any chance we can see a video of the mold making and casting process? I spent a month in the foundry during my apprenticeship (probably 1967 or 1968) but we didn't go in the pattern shop. So while we obviously used split patterns like this, it is long enough ago that it has become a bit sketchy. I imagine there is no shortage of casting channels on TH-cam, but that won't be the same as following this part through the process. A great video Keith.
love the CAD drawing.. (Cartoon Aided Design) (with apologies to either Tom Lipton or Keith Fenner, where I heard that first used) GRIN
Nicely done and presented very well. Very elaborate glue up on the pattern. Did you mention this was a museum? Are you a volunteer or a member of the staff. Either way, thanks for taking the time to make these videos.
You never fail to amaze!!!!!!!
What did I do?
You think at a level soooooooo very far above the norm. You and both your daughters. It's amazing to share the planet with you.
Kary Miller
Thanks - you're being silly now. Our comments are showing up on this guys TH-cam video. I don't know how this Google+ thing works. My original comment was about this fellows work.
Hahahaha, bet he wonders what the heck!!!
A slightly different way to register the two halves is to first glue them together with paper between them. Then drill a holes that go through one half and into the other. Drive your wooden dowels into the holes and now the two pieces will be perfectly aligned when you remove them. This way you don't have to worry about the two pieces rocking slightly off axis when you drill the holes in each half.
If you use paper between the two half of the pattern when you do the glue up you will have a strong enough bond so the halves do not split when turning but you can then insert a knife at the paper and actually split the paper.
I think that tubalcain has a video showing this process.
Congratulations. I just watched Keith turn these pulleys for you.
Thanks for another great informative vid. I have two questions.
Why not weld the pulley instead of replacing? Old machines are often found with repairs, so you would still remain true to the technology and save hours of work. But then again, we wouldn't see how this process works so thanks for the extra effort.
I've never turned in wood, but instead of using all these segments, wasting hours in the process, why not use MDF to make up your halves. MDF is very stable to temp and humidity, extremely flat, glues well, and is easy to work on high speed shapers. Plus there is no end grain and no internal segment joint seams. Does it not work well in a wood lathe? That said, I appreciate seeing the old methods using lumber.
*****
Thanks
Wondering.... Do you compensate the size of the pattern for the shrinkage of the cooling metal?
Yes, I used a shrink rule to do this - which takes in the shrinkage factor for cast iron.
Oh, okay I know as I was an old fabricator heat and shrinkage were a factor always needed to be taken in mind. Thanks Keith
@@VintageMachinery Hey Keith, you did use a shrink rule. For the record, shrink rules are not calibrated for any specific metal. Rather they adjust for a specific amount of shrink. For example 1/8 in per foot. In theory a strip of iron will shrink about 1/8 in per foot. In practice that’s not always the case. Depending on the iron and the shape of the pattern, you may not see any shrinkage. The core serves two purposes. First it gives you a rough starting point for machining. Second it evens out the thickness of the casting features resulting in more even linear cooling and a reduced dependency on risers to mitigate shrink voids in heavier sections. Thirdly by making the cooling more even or linear, it greatly reduces internal stresses in the part.
If you were concerned about keeping the "drive" end stronger, and left the 1 inch core oversized to 3 inches to allow this, why not turn the pattern 180 degrees and have the thin core on the centre end, allowing a drive though the 3 inch core?
When using the joiner, the leading edge almost always ends up thinner than the trailing edge. I like to turn the piece end to end each pass to keep things even if the grain allows.
In theory this is false. In practice this is true. I noticed this abnormally on my porter 300. In terms of blow out. I am not a fan of fancy insert heads. This may be one case for them, I don’t know. But the jointer is about accuracy. To many folks forget this chasing surface finish from a jointer. A scraper plane such as a Stanley 85 would truly solve any surface issues here. A Stanley 112 would also work and it had an adjustable angle.
I'd have liked to have seen a small faceplate \ flange used at the tailstock end.
Graet videwo
Gouging always worries me, with the guide being set a half inch or so away from the piece being worked plus the increasing distance away from the guide when turning, my self preservation says that the Gouging tool could catch on a Knot or a fixing screw and flip/catapult/ seesaw of the guide and take your head off, I would have thought that a second guide situated above and in front/ (closer to the operator) of the bottom guide would be 100% safer, that said I am aware that the method here has been used for years so are my concerns unfounded ???
***** Point taken Keith, experience is everything, but it seems to me that there's a heck of a lot of power/velocity which could catch an operator off guard in a split second if your concentration is distracted
I found it rather amusing that in part 1 of this project you fussed with the miter trimmer to make all 6 segments fit perfectly with no gap. Now you still went ahead and jointed the mating surfaces flat. So It was totally unnecessary to be that particular.
You gotta love wood glue ups. The parting line is one thing. But the segment joints for the most part were tight where thru count.
3 D printing is more efficient nowdays
***** That s great, if you want to preserve the technology. Unfortunately in factories you don't work like this anymore. You use CNC or 3 D printing. In Hungary for example at Opel factory the prototype for a gear box in made with a 3 D printer. Using an older method with wood would take about 6 months. Wiht CNC is very hard to do it, 3 D is the ideal solution. It takes only 6 weeks do finish it.
Szaby Sascha 3D printing is a fad. It replaces the woodworking skill sets so an idiot can do it. Your efficiency is achieved by placing the design into solidworks or whatever your using. I machine wood all the time on metal working machines and use CNC control from time to time to do it in making infill hand planes. In terms of getting there... your either going to remove material or add material. Two different viewpoints... same end result. I am more interested in the thermodynamics of cooling metal. How does the metal flow in the mold. How does it shrink. Do I need a match plate. etc. etc. 3D printing can help you but its not as effective as you think. Your still going to need to make the final parts. Most plastic injection molds today are done in solidworks, emailed to china, and the final injection dies are fedexed back to the US. No 3D printing. Just CNC machines removing material. For what we do, this method still works quite well. Taking an old time pattern maker six months to make a pattern strikes me as strange. Are you guys training the patternmaker during that six months? It took MoPar two weeks to build a new version of the 426 HEMI engine after a significant design change. (Set of wooden chrysler patterns).
I don't know the 6 six months thing. I just read it in a technical magazine from Hungary. An engineer wrote an article about this topic, and he said that. Here in Romania there isn't any training for this profession any more. My dad learned this in the 60's , in 2005 the factory where he worked said that they don't need this department anymore.
Sadly, he is correct. The pattern world has been replaced by Solidworks which is an incredible software program. But wood was cheap and quick to machine so it was a perfect medium for patterns and prototypes. Now, we have CNC machines that can cut solid aluminium faster than slick willy. Just dont let the 3D printer fad catch you up. In some cases, its great. Usually in those cases, the reason the 3D printer can do the job is precisely the reason the foundry guys are going crazy. Patterns still have to be rammed up on a squeezer line and that means core prints, cores and draft. Some things change. Some things will never change.
You re right Arnold.