3D Scanning for posterity

[AURktman]

While watching / participating in the chrome shop topic, I noticed something that I have thought about before regarding some of the more rare / priceless items that go with the rare cars.  The comments were about rare parts being ruined during the chroming process.  I know many details can be wiped out with unskilled chrome shops, but there is a way to mitigate (not solve) these issues.  If you have the item 3D scanned BEFORE you send it out for chroming, it is possible to take that file and have the part / item additively manufactured or CNC milled if a replacement is needed.  Plus it also saves that part for posterity if and when the parts are no longer available.  

 

Scenario:  You put all of the chrome items back on your 1939 Lagonda LG6 Rapide Drophead and take it to the Las Vegas Concours d’Elegance (just work with me here on this, not saying this is remotely possible, feasible, or even likely).  At the show, some drunken idiot takes off in a golf cart and runs amok and smashes into your Lagonda.  After the crowd pummels the offender into unconsciousness using champagne bottles and high heels, you notice the rear view mirror mount and marker light is smashed beyond repair.  

 

After looking on RockAuto and AutoZone websites, you determine its not possible to find another. Well, luckily you had the parts 3D scanned by a local community college who has an additive manufacturing program with scanning capability.  You take that file to a high-end CNC shop and they mill another out of your choice of stainless, brass so you can plate it in chrome, or aluminum.  

 

Now, original parts are ALWAYS preferable, but if there are no replacements available then you are left without options.  Here are examples of some mid-range quality scans and videos I've found on the web. In the end, I know and appreciate the scarcity of some of the components on these vehicles.  I fully embrace using original parts where you can as well, but if you're left without options OR left with only one in existence it may be wise to get it scanned. 

 

 

 

 

 

Edited January 10, 2021 by AURktman (see edit history)

[Flivverking]

I never saw any normal reason not to use well reproduced parts when the original part is in a practical sense  unobtainable in a reasonable amount of time or extinct (or nearly so)and or priced way out of budget.

 

I personal wouldn't worry about pre scaning parts for the future unless you have a market and are going in the business to reproduce the parts for a minor few .

Where does it end?

Why not blue print and scan the whole car ? 

 

It's not like 1939 Lagonda LG6s  are rare..I have two for parts !LOL😜

 

 

 

 

 

[AURktman]

Well, that is kind of what I was saying.  No one is suggesting scanning parts for a 1931 Ford roadster, there are thousands of those and you can find a bunch of OEM/repop parts everywhere.  The ones I'm talking about are the 1 of 5 or even 1 of 50 types of car.  As for scanning the whole car, I think that is a really good idea for the rare ones.  

[Terry Harper]

Well that would be great. 3D scanning is used in industry for comparison and quality assurance (new manufactured part against specs.) and of course for reverse engineering.

 

The reality is affordability and having the technical skills. I love these videos that make it seem so simple.

 

A good 3D scanner has been on my wish list for my classroom/lab for quite a few years. Unfortunately quality and capability comes at a price - near $60,000.00 for a Faro unit. Then there is the CAD software such as Solidworks - to manipulate the 3D models. We have an educational license but a commercial license is approx. $8,000.00 per year plus another approx. $2,000.00 annual maintenance fee. Quite expensive especially if your just using it for hobby purposes. Then technical skills to use the software. What they don't talk about is the time to work through the 3D model to correct any imperfections and to assign fits and calculate tolerances critical to manufacturing the part and maintaining design intent. For instance lets say the scanned model indicates a measurement of .50012" between milled faces. Was that the intended dimension? Most likely the nominal dimension was .500" so the scanned model needs to be modified and a tolerance determined based on the type and class of fit required to allow it to slide, rotate or be located to or with another part... Without this critical step the resulting CNC produced part may not function or fit. All this has to be determined by someone with technical skills and experience.  If you have to hire all this out its not cheap.

 

Here is an example:

Recently we complete a live work project for a non-profit. The project involved fabricating a radiator cap. One of my students set to work reverse engineering the cap and locking knob using old photographs and measurements of the filler neck. Then they developed 3D models in Solidworks as well as the 2D shop drawings. They also 3D printed a mock-up so we could test the fit etc. Our original intent was to have the cover and knob cast but COVID played havoc with that so the decision was made to mill it out of solid aluminum and replace it with a cast iron version later. We used a CAM program to develop the tool paths and post process to G-code. Once that was done we did a test run in machinable wax. Then ran the final part. As it turned out the locking knob was 3D printed in metal for us.

 

With private or commercial clients we provide them with a estimate of the value of the work and ask for a donation (partial or full according to their generosity) and the cost of materials. Since this particular client is a non-profit who we have done a large number of collaborative projects with and the material was donated there was no charge. However, just out of interest we calculated the cost if we were charging for time and materials:

 

Design services:
CAD: 4 hours @ $45.00 = $180.00
Engineering: 10 hrs. (includes field visit/mtg.) @ 120.00 = $1,200.00
(Design/drawing review, calculations, mtg. w/ design & mfg. team: 2 hrs.
(Field visit: 6 hour round trip, 2 hours on-site.)

Manufacturing services:
CAM (generating setups, tool paths, simulation, post process to G-code) 6 hours @ 45.00= $270.00
CNC Machine time plus operator: $1,500.00 (includes test run with machinable wax)
3D Printing (PLA Mockup): $89.00
Material: Alum, machinable wax, coolant: $75.00

Total: $3,314.00
Total does not include cost of 3D metal print for the knob)

 

Granted, without a site visit the cost would have been lowered by $960.00 however, what if the client got the measurements wrong? Or the photos missed a critical detail? Who would be liable for the additional costs??

 

My point is that people seem to think that 3D scanning, rapid prototyping etc. can produce cheap parts at the snap of a finger. It reminds me of the many times when I worked in architecture and handed a client two or three drawings and a bill for $5,000.00 Looking at those few thin pieces of paper they simply could not visualize the justification for the time, technical skill, professional knowledge and liability that those pieces of paper represented. Same with 3D technology. Its an amazing tool, I love working with it and it has opened up a whole world of possibilities but.... it comes at a cost that people must be willing to pay.

 

 

 

 

 

Edited January 11, 2021 by Terry Harper (see edit history)

[BearsFan315]

WE use 3d scanning for QA as well as validation.

 

also we use it to scan our old WOOD tooling for castings. many of them are over 50 years old. we have them scanned in then we can modify/manipulate the scanned model to fix the wear and any updates required to the new forms/molds to be made. when we first started doing this years ago we did it the NEW way which was the Faro arm or Roamer Arm (competing companies doing the same thing) was a long drawn out tedious process. arm one side where ever you could reach with the arm, then flip the part, reset, manually line up and do that side. was really hard to get into detailed area, and any blemishes. lot of user input and interaction. then along came laser scanning. you put on a few Targets and then can scan the entire part in minutes. flip it all around scan all sides, inside out, or wherever the scanner can scan. the Targets keep the part aligned in the system. 

 

the only drawback to scanning, is the desired resolution. some of the files are GB in size !! depending on the clarity and detail you want. you can scan and see scratches, dings, etc.. on some cast parts you can see all the deviations and the mold/form marks. 

 

we also reversed engineered one off parts we produced that there were no drawings or patterns for. 

 

i spend many hours reviewing scanned images, comparisons, deviations, etc...  

 

ne thing is you can convert and send to a 3d printer and make a quick prototype, or send to a tooler and have a tool made from the model and go from there.

 

i know on my 1929 Chevrolet i have digitized several parts to have them fabricated. most recent was the molds to make the one year only pedal seals !!  

[JV Puleo]

Terry, you have reminded me of a conversation I had with another machinist - actually the gentleman I bought my B&S mill from. He runs a successful small machine shop in the Philadelphia area equipped with many older  Brown & Sharpe machines. At the time we spoke he did not have any CNC equipment. Why? Partly because the overhead costs involved made it necessary to run the equipment almost constantly. As a result, there was intense competition between shops, most of which were over-invested in the latest and greatest machines. In order to keep those machines running all the time they cut prices - so, he could always find a CNC shop that would do a job for him for less than he could do it himself...but, most of the time, this wasn't necessary. He related a story about one job...it was a short run (perhaps 25) sets of crankcases for an early Indian motorcycle. His job was to machine the cases. When asked if he had CNC equipment he said no. "How can you do it then?" he was asked. "How do you think they did it in 1913?" was his answer.

[Terry Harper]

11 hours ago, JV Puleo said:

When asked if he had CNC equipment he said no. "How can you do it then?" he was asked. "How do you think they did it in 1913?" was his answer.

I like that one Joe!

 

There is a time and place for everything.

 

Here is a good example:

A while back we were asked if we could replicate some hinges for a set of Vesta headlights. After reverse engineering and modeling in Solidworks We developed and 3D printed fixtures and even ran a test run in machinable wax. Granted, we did this more to see if we could do it than anything else (I am always up to a good challenge!)

 

In reality, using CNC to produce these parts is prohibitive in terms of time, and material so we 3D printed patterns and these will go to the foundry. We will then use the 3D printed fixtures for milling the slots and drilling the holes. I have not tallied the time on this project. Not including the time spent running CAM and the CNC test run I don't think the cost would be all that bad.

 

In the photo below the green pieces are the 3D printed patterns (the still need to be cleaned-up and painted) the blue assembly is the CNC test run in machinable wax.

 

 

Edited January 11, 2021 by Terry Harper (see edit history)

[Restorer32]

Why not just use the original pieces as foundry patterns?  Shrink would be negligible on such a small part.  We do things like this regularly with no problems. Those 2 pieces in silicon bronze might cost less than $15 each. A bit of machining and you're done.  For a bit more money, but still less than the cost of having them 3-D printed, they could be lost wax cast, which would require little if any machining.  Just being the Devil's advocate here.  As the president of one of the largest 3-D printing firms said, 3-D printing is primarily useful in making parts that cannot be made in any other way. 

[Terry Harper]

18 minutes ago, Restorer32 said:

Why not just use the original pieces as foundry patterns?  Shrink would be negligible on such a small part.  We do things like this regularly with no problems. Those 2 pieces in silicon bronze might cost less than $15 each. A bit of machining and you're done.  For a bit more money, but still less than the cost of having them 3-D printed, they could be lost wax cast, which would require little if any machining.  Just being the Devil's advocate here.  As the president of one of the largest 3-D printing firms said, 3-D printing is primarily useful in making parts that cannot be made in any other way. 

 

True, we could have used them as the patterns. However, in this case the client wasn't very comfortable doing that. (given the current state of the mail system I can understand) Since we do all our 3D printing inhouse its not a big issue. Also the client has requested a number of sets - we can print multiple copies of the pattern so the foundry can do one flask and be done. We are also doing the latch as well. The original looks to be pot metal and is incomplete and in pieces so no choice but to 3D print the pattern with that one.

 

The funny thing about that statement is all the parts we deal with in this hobby were indeed made in another way.

 

 

 

 

[Restorer32]

Amazing technology and I'm sure it has a place but I would like someone to come on here and give us the actual cost to have a one off piece made. Not "My nephew does CAD so he did it as a fun project for free" and not a "guesstimate" but the actual cost if one had to pay professionals the full cost from scanning to a useable part. Then we would have something to actually compare.

[Terry Harper]

Quote

the actual cost if one had to pay professionals the full cost from scanning to a useable part. Then we would have something to actually compare.

 

That would actually be of little use since projects vary widely in complexity and size and deliverables. Its akin to comparing apples and oranges. What I can do is modify the cost for the radiator cover (which was indeed submitted to the client for tax purposes so its not a "guesstimate") to include scanning and back out the site visit. I have marked the line items that have changed. As for "My nephew does CAD so he did it as a fun project for free"  I have 30 years working in design/engineering with 15 of those years as a consultant to backup 8 years of teaching this stuff.

 

Design services:

*Scanning: $300.00 (most firms charge a minimum fee per part which increases based on size and complexity)
*CAD: 2 hours @ $45.00 = $90.00
*Engineering: 2 hrs. (includes field visit/mtg.) @ 120.00 = $240.00
(Design/drawing review, calculations)

Manufacturing services:
CAM (generating setups, tool paths, simulation, post process to G-code) 6 hours @ 45.00= $270.00
CNC Machine time plus operator: $1,500.00 (includes test run with machinable wax)
3D Printing (PLA Mockup): $89.00 (Many 3D printing firms charge by the weight of printing material used (cost plus) so again, size and complexity can vary the price widely)
Material: Alum, machinable wax, coolant: $75.00

Total: $2,564.00
 

If you do a bit of research you will find that some of the hourly rates quoted above are actually cheap but we are also an educational institution as opposed to doing this for a living. Again, the cost can vary widely based on the size and complexity and the clients expectations.

 

As with anything else there are many people offering services to scan etc. The question is what exactly are they providing and what is your expectation versus their experience and capabilities? For instance they may only scan the part and provide you with a rough file while what you really need is a file that is ready to go to CNC or ready to 3D print as a foundry pattern with draft, shrinkage and machining allowances.... big difference!

 

 

 

 

 

Edited January 11, 2021 by Terry Harper (see edit history)

[Gary_Ash]

I've been working with Edinmass to help make a new impeller for his waterpump in the 1917 White.  Ed sent me the old cast aluminum impeller - pitted, corroded, and broken - and I made a 3D model from that just using digital calipers, no scanning.  I added on some extra material for finish machining and scaled it up 2% for casting shrinkage in bronze.  I printed out four patterns in PLA, sent these to an art foundry where they are being cast in silicon bronze this week using a "lost PLA" process.  After machining by JV Pulveo, Ed will have his new impeller and a spare, and the others (not machined) will go to homes at museums.  The CAD work, 3D printing, and machining are being done "pro bono", haven't got a final cost on the actual casting.  No one would want to pay me or Joe what we would charge for our time, but Ed has offered some quid pro quo that satisfies.  Each 3D print took 9 hours, but the machine runs by itself all night.

 

I did send the CAD file to an on-line machine shop to see what CNC machining would cost by comparison.  For making four 3.5" diameter impellers, the machining cost was $183 each in 6061 aluminum and $532 each in bronze.  So the cost for "4-off", as opposed to 1-off, doesn't seem too bad if you want machined aluminum parts.  As Terry indicated above, there is considerable effort to go from a CAD file to the CNC machining, even for one part.  I guess the bronze machining goes more slowly.

 

Four 3D-printed patterns in PLA plastic and the original aluminum impeller.

 

As for scanning and 3D modeling an entire car "just in case", that's probably a lot of wasted effort, sort of like building the Maginot line in France before WWII.  For reproducing sheet metal, Wray Schelin's "flexible shape patterns" can be used to easily copy a fender, hood, or other body panel.  If a right fender is damaged beyond repair, a flexible shape pattern from the left side can be turned inside out to create a new right fender or a pattern can be made from a different car.  No scanning or CAD work required, no wooden body bucks!  Here's how they work to make a new fender using a mallet, stump, and English wheel (long videos):

 

 

[padgett]

Old IBM joke: $2.50 for hitting with hammer, $2500.00 for knowing where to hit.

[Terry Harper]

4 minutes ago, padgett said:

Old IBM joke: $2.50 for hitting with hammer, $2500.00 for knowing where to hit.

 

Absolutely!

 

Hitting it in the right spot $2,502.50

Hitting it in wrong spot........  $$$$$$$$$$$$$$$$$$

 

As we found out when someone "hit" our large format copier/scanner in the wrong spot..... $14,000.00 later.....

[MikeC5]

Gary, that's a really good example of something that can be adequately modeled just by taking dimensions with a caliper.   The 3D scanning is really only necessary if a part is complex with 3D curves, etc.  If someone knew they were going to do this many times, it might be worth investing in a copy of AutoCad and learning how to use it.  

[SC38dls]

Gary I hope you held out for a good ride in the Great White on top of the quid pro quo !!! 

[edinmass]

3 hours ago, SC38DLS said:

Gary I hope you held out for a good ride in the Great White on top of the quid pro quo !!! 

I was thinking of a drag race.........White VS Stude............🤔

[SC38dls]

Ed be careful you could be handing over the pink slip for that Great White!  Studebaker’s had a lot of speed records back in the day. I even saw a 289 Studebaker crush a Mustang 289 owned by the younger brother of the Ford company NASCAR diver in a quarter mile drag at a track. 

dave s 

38 Studebaker State Commander proud daily driver owner !  

[padgett]

"289 Studebaker crush a Mustang" R2 ?

[1912Staver]

On ‎1‎/‎10‎/‎2021 at 4:46 PM, Terry Harper said:

Well that would be great. 3D scanning is used in industry for comparison and quality assurance (new manufactured part against specs.) and of course for reverse engineering.

 

The reality is affordability and having the technical skills. I love these videos that make it seem so simple.

 

A good 3D scanner has been on my wish list for my classroom/lab for quite a few years. Unfortunately quality and capability comes at a price - near $60,000.00 for a Faro unit. Then there is the CAD software such as Solidworks - to manipulate the 3D models. We have an educational license but a commercial license is approx. $8,000.00 per year plus another approx. $2,000.00 annual maintenance fee. Quite expensive especially if your just using it for hobby purposes. Then technical skills to use the software. What they don't talk about is the time to work through the 3D model to correct any imperfections and to assign fits and calculate tolerances critical to manufacturing the part and maintaining design intent. For instance lets say the scanned model indicates a measurement of .50012" between milled faces. Was that the intended dimension? Most likely the nominal dimension was .500" so the scanned model needs to be modified and a tolerance determined based on the type and class of fit required to allow it to slide, rotate or be located to or with another part... Without this critical step the resulting CNC produced part may not function or fit. All this has to be determined by someone with technical skills and experience.  If you have to hire all this out its not cheap.

 

Here is an example:

Recently we complete a live work project for a non-profit. The project involved fabricating a radiator cap. One of my students set to work reverse engineering the cap and locking knob using old photographs and measurements of the filler neck. Then they developed 3D models in Solidworks as well as the 2D shop drawings. They also 3D printed a mock-up so we could test the fit etc. Our original intent was to have the cover and knob cast but COVID played havoc with that so the decision was made to mill it out of solid aluminum and replace it with a cast iron version later. We used a CAM program to develop the tool paths and post process to G-code. Once that was done we did a test run in machinable wax. Then ran the final part. As it turned out the locking knob was 3D printed in metal for us.

 

With private or commercial clients we provide them with a estimate of the value of the work and ask for a donation (partial or full according to their generosity) and the cost of materials. Since this particular client is a non-profit who we have done a large number of collaborative projects with and the material was donated there was no charge. However, just out of interest we calculated the cost if we were charging for time and materials:

 

Design services:
CAD: 4 hours @ $45.00 = $180.00
Engineering: 10 hrs. (includes field visit/mtg.) @ 120.00 = $1,200.00
(Design/drawing review, calculations, mtg. w/ design & mfg. team: 2 hrs.
(Field visit: 6 hour round trip, 2 hours on-site.)

Manufacturing services:
CAM (generating setups, tool paths, simulation, post process to G-code) 6 hours @ 45.00= $270.00
CNC Machine time plus operator: $1,500.00 (includes test run with machinable wax)
3D Printing (PLA Mockup): $89.00
Material: Alum, machinable wax, coolant: $75.00

Total: $3,314.00
Total does not include cost of 3D metal print for the knob)

 

Granted, without a site visit the cost would have been lowered by $960.00 however, what if the client got the measurements wrong? Or the photos missed a critical detail? Who would be liable for the additional costs??

 

My point is that people seem to think that 3D scanning, rapid prototyping etc. can produce cheap parts at the snap of a finger. It reminds me of the many times when I worked in architecture and handed a client two or three drawings and a bill for $5,000.00 Looking at those few thin pieces of paper they simply could not visualize the justification for the time, technical skill, professional knowledge and liability that those pieces of paper represented. Same with 3D technology. Its an amazing tool, I love working with it and it has opened up a whole world of possibilities but.... it comes at a cost that people must be willing to pay.

 

 

 

 

 

 

 

 

I think this is a good example of the limits of this technology as applied to the old car hobby. The final part is nice but the cost is out of reach to most old car owners. It would be a pretty simple pattern to make by hand for anyone with even average wood working ability's. And not a particularly " final size " sensitive part, meaning a very precise shrinkage allowance is unnecessary.

  For a great example of what is possible by a reasonably talented , old vehicle hobbyist . Using a modest workshop . Look no further than " Great War Truck" on the HMVF England site.

The results are quite amazing.

For people that have access to high tech equipment as part of their day job, why not. But as a solution for average old vehicle owners the costs as I see them are just not justifiable in any but the most extreme cases.

 

Greg

 

[Terry Harper]

2 minutes ago, 1912Staver said:

 

 

 

For people that have access to high tech equipment as part of their day job, why not. But as a solution for average old vehicle owners the costs as I see them are just not justifiable in any but the most extreme cases.

 

Greg

 

I totally agree Greg. There is a time that it makes sense and when it doesn't. In this case using our CNC milling machine to mill it out of solid was over the top and I would have really questioned the sanity of any shop proposing to fabricate it that way. In this case we did it just to see if we could do it and to push our experience with the technology. As I mentioned, we 3D printed the foundry patterns but for the moment our plans with the foundry went sideways so we got thinking which can lead to interesting things.

 

As Gary pointed out. A lot of good work can be done with a caliper. However, if you have to hire someone to use the caliper or carve the pattern than expect to pay. I guess my point - which I struggled to make clear, is that the technology is amazing but its not a magic bullet that's going to solve all or solve it cheaper. That there are appropriate applications and inappropriate applications and that it cost money  - especially if you don't have access to it or can do the work yourself. In the wrong application it can be a very expensive experience indeed. 

[BearsFan315]

as i was told growing up there is an A@# for every seat

 

and in college was told there is more than one way to get to a solution, have to find the one that works for you (cost, time, effort) what skills do you have that can get you to where you need to be ? some are good with wood, some with metal, some with both, some with none. i can barely get by with either, but i can draw/design most anything in 2d or 3d, paper or cad. and bought a cheap small 3d printer years back and have done may projects on it for my car, things around house and many prototypes for others. 

[edinmass]

4 hours ago, SC38DLS said:

Ed be careful you could be handing over the pink slip for that Great White!  Studebaker’s had a lot of speed records back in the day. I even saw a 289 Studebaker crush a Mustang 289 owned by the younger brother of the Ford company NASCAR diver in a quarter mile drag at a track. 

dave s 

38 Studebaker State Commander proud daily driver owner !  

 

 

Wasn't planning on racing for pink slips......or money. Just the fun of it. I'm quite sure the only way the White would pass the Stude is if it were inside an airplane.

[Gary_Ash]

Now's your chance, Ed - my engine doesn't run yet.  Of course, if you don't have a waterpump operating, we'll both be sitting on the starting line.

[A Woolf]

For those you might be interested there is an inexpensive way to acquire a copy of Solidworks.  The Experimental Aircraft Association (EAA) has a deal with Dassault and EAA members can get a copy of the software for the price of membership.  I think the membership is about $40.00 year.    This is a full featured copy of Solidworks and the only restriction is that it for personal use only.   Plus you get a neat magazine several times a year.  

 

I spent a lot of years in the aerospace industry and have an extensive background in 3D modeling and have  used solid modeling a great deal in my current restoration work.   Solidworks is a very good tool and I think the EAA offer is great deal.  

 

Alan