Project Kevin Part 2: The Kevining

Project Kevin is complete (for now, we have to invent a body and that might be... interesting)! The last time I updated this, I had left you at "We dug stuff up and were making it as pretty as it could be" in the lab. That left us with a pile of neat looking bones, but of course we wanted more. Did we have enough to make a skull? And if so, what did it look like? We sure thought we did based on what we identified, though since the site was trampled int he Cretaceous most bones were missing chunks.

One of the mostly complete brow horns

First things first though. Let's laser scan (alright not a laser, but an Artec Spider structured light scanner) all the things! This gives us a good baseline to record what we have. These scans can also be shared with interested researchers across the planet. Researchers are usually pretty happy giving opinions of things and many helped us with details on how this thing might come back together.

Right maxilla in digital form

We can also try out new things with the scans. I came up with an interesting idea to print a 30% scale model of all the parts (using mirrored parts if one side was missing or just incomplete/really ugly). We popped off the parts on our Formlabs Form2 (the 30% scale was determined by the size of the build plate of the printer, these SLA printers can be pretty small) and tried to put together a model to guide us. We goofed though.

The first draft, complete with all our errors

Turns out we had the brow horns on backwards. Also the fits between the bones weren't as accurate as we would have liked. So we fixed them on a small scale before committing on the casts and prints of the full scale stuff. It also gave us the ability to try out things like a scaled and computationally-squished rostrum to make a part that we were completely missing. In this case we printed out a bunch of different possibilities and fit them on until we had a result that looked plausible.

Second draft of scale model, now we get to try out different beaks

With this information in hand, 4 binders of papers as references and a pile of casts, we were ready to take a stab at reconstructing the skull. Lainie and Grace really did a heck of a job learning these techniques. Printing out full size mirrored parts make the skull more accurate and easier to reassemble than if we were to sculpt the missing bits from scratch.

3D prints, casts, lots of epoxy putty and Bondo. Lainie for scale

Things went pretty smoothly till someone (who shall remain nameless) suggested our minimum length conservative frill was probably much too short. The first draft was based only on the length of the frill parts that were preserved. Chasmosaurines like this have seriously long frills though so we took their advice and busted out the sawzall. It was only plastic after all.

OK, let's move this frill about a foot to the back I really think it looks better this way.

We also had to make teeth. Hydrospan 100 was wonderful for this. We poured it into a mold of Ava (RMDRC 12-020)'s dentary and made a floppy cast. This material was then soaked in water until it expanded enough to fit the tooth rows for Kevin. Then we molded it, poured a bunch of plastic copies and played dinosaur dentist for a few days getting over 100 rows of teeth in all the jaws.

We molded all the finished parts and made casts. The skull was cut apart to make the molding process easier. A single mold for the top of the skull would have been huge, complicated and really heavy.

This is what a Kevin skull kit would contain if you bought one

Grace and Lainie making a huge mold for the frill

We put the prototype together in just a few days. It was a lot bigger than we expected. But after paint and finishing, I thought it went together pretty well! Technology really helped us out on this one, saving us materials, time and most importantly effort. I don't think we could have gotten it done by the deadline without it. Now it's time to figure out where this belongs in the family tree.

The prototype is done!

And the obligatory "Curator for scale" photo. I forgot to suck in my gut.

Here's what we put together for the left side, and where we got it from.

Scanning a Daspletosaurus

So we all know that fossil skeletons are rarely found 100% intact. This is especially true for larger animals for a number of reasons, from scavenging to paleoerosion to roots to modern weathering. Even with the mighty tyrannosaurs this is true: Sue? Missing bits. Ivan? even less complete. Sir William? The dozens of shed tyrannosaur teeth at the site might be a hint where lots of the parts went. Yum.

Field map of quarry

Quarry during dig

During excavation of Pete III, the TPI field crews kept careful records to document the site properly (as all paleontologists should, I don't care if you're academic or commercial). As we excavated and uncovered bones, they were assigned field inventory numbers, recorded and were mapped to scale. It quickly became apparent that there were only enough leg elements for one. The other one was gone, but where did it end up?

Ankle bone mirrored in the computer

Microstratigraphy along with the mapping provided a possible answer. The sediment indicated that Pete III was located in a crevasse splay deposit. These deposits are formed when a natural levee breaks, squirting water and sediment over the adjacent floodplain. It's like a geological zit. This particular rupture apparently contained a Daspletosaurus in early stages of decomposition, where the bones were still articulated and loosely held together, though some were washed further on. Unfortunately, we think that's where the right leg ended up. We were 75 million years too late.

Original and printed bones, various colors of PLA

OK, so I had to sculpt one bone by hand

So what can we do? A one legged Daspletosaurus mount looks stupid. In the old days we would sit down with some material, a ruler and our eyeballs and sculpt the mirror image by hand. Luckily we don't live in the stone age anymore and instead we have computers and lasers and hot plastic.

Project layout. Pete's skull will be about 1.8x the size of the cast on the table.

We were missing one phalanx completely, so we laser scanned one from our Albertosaurus cast, blew it up to size, and then printed it. The entire right foot was scanned from the left, mirror imaged in the computer and then printed in full scale. We're still in the process of scanning the femur, tibia and fibula, but instead of printing them (which would take a lot of time) we are sending the files to someone with a 5 axis CNC router to carve them for us. Work smarter, not harder!

3d scanner and printer up and running

After several years of fighting with technology, our 3d laser scanning program is up and running here at the RMDRC. We're currently scanning original specimens that are in the restoration queue in order to generate mirror image parts for missing bones. We started with simple parts on the Avaceratops including cranial and limb material, and progressed to more complicated things like the dentary and maxilla. Not satisfied with the amount of punishment we dished out on ourselves, we then tried Enchodus bits like the super thin premaxilla and operculum. They all turned out extremely well.

One of our new printers which works surprisingly well

Now that we have more complete sets of replicas to work with, our restoration of these specimens will be not only much more accurate but also faster.

Avaceratops lower jaw, with white 3d prints of actual material

Avaceratops tibia, split in a computer so it could fit in the printer

Next on the docket is some of the cranial material of our Daspletosaurus specimens. Donor parts from other animals will be brought in and scaled to size. Surprisingly easy to do with the software and operators we now have here. If you are out there and need some items scanned, give us a call!