07/28/15 It's been a month since Pikes Peak and lots of things have happened in that time. I've been lagging in blog posts because it's enough work just making the things happen. I did get a few videos rendered and uploaded. Here's the multi-camera one, just search for the others.
The Drakan Spyder production continues - we actually delivered the first (blue) car back in early June, before the race. Shinoo and Joe from Sector111 came up then to help complete the build and install the drivetrain. Now roller chassis #2 (red frame) is nearing completion and is scheduled to leave the shop Friday. After this one, 8 more to go in the current batch.
A couple of weeks ago we had another outing at ORP. It was the first time that one of each model we currently make was on the track at the same time. Speaking of ORP, those wanting to check it out in rFactor2 can now download the public beta.
We got to test the D2RS with the boost controller unplugged, and then plugged in. Several useful things learned there. First, it confirmed that we would have done as well or better with just a regular D2 - lap time without boost was essentially identical to our demo D2 on same tires. We also confirmed the correlation or ORP lap times to Pikes Peak times which will be extremely valuable in testing for future events. With boost connected - yes, the car was several seconds faster but at sea level the extra power was overwhelming the well-used rear tires in less than a lap. We'll need a fresh set and probably in different compound to really push it. That'll happen at some point and we know the car is capable of overall record here, in street legal form including emissions.
Jonathan trailered his D4 up from California and Neil brought his D1 on the newly built trailer (several good upgrades from the first one, we always keep evolving!). Dedicated track cars need not be inconvenient.
It was gratifying for me to look in the paddock - a lot of my 'children' gathered at once. I've come a ways from the early napkin sketches :)
We've run into some technical roadblocks with the D2 restyle at Lancair, so now the plugs are being made at Willamette Pattern next door - the company that cut the D1 plug way back when. They've acquired new capabilities since then and collectively we've learned a few things so it's all being applied. The work is coming along nicely. One plug is done and next is well on its way.
There are many other developments and I'll try to be more timely with the posts. Stay tuned.
8/19/15 Some good progress on plugs for the new D2 bodywork - the guys at Willamette Pattern are cranking them out! There are some unique methods we're trying that are working quite well. This stuff is still very much an art form even with all the technology, and hand finishing is still very much a part of the process (trying to minimize that though). The foam patterns are coming out perfectly and we're now applying and sanding several coats of Duratec primer.
In other D2-related news, production of the Sector111 Drakan cars continues with two now delivered and several more on the way.
Several other things are in progress as well. With that, our demo D2 is now for sale, to make room for the next development car. Asking $65K - email me for details.
08/26/15 We're pushing to get the next few Drakans out. So a lot of machining and assembly... Chassis #6 is welded up.
Front and rear bellcranks are machined for the next few cars.
Chassis #4 is being assembled with customer-supplied drivetrain.
In the meantime new D2 patterns are moving ahead. The pictures below show just how much bondo and sanding it takes, even with CNC machined tooling foam. Since the guys are busy with other projects I'm doing a good deal of the sanding myself (Peter is also doing a lot of it).
09/24/15 Almost a month since the last D2 update. The majority of the effort during this time has been focused on building the first 5 of the 10 Drakans we have on order, and now the Stiletto is picking back up again - this time for the final push to completion. There is some upcoming press coverage on the Drakan, look for it on the web and I'll post links when I get them.
Three Drakans are done, two more are progressing rapidly. There are also several things that Drakan-specific development has produced, driven by requests from Sector 111, that has found its way back on the D2s. Reduced-caster front suspension is one, dead pedal is another. The latter is more challenging than might seem, since the pedals have 5" of adjustment range and the dead pedal has to travel with the rest. Below is a prototype that works pretty well.
But there is some D2 specific stuff going on as well. Most of it revolves around the new bodywork. First, the bodywork itself - another plug is now machined (one more to go for roadster version, then coupe-specific parts later). Despite CNC machining there is still a lot of hand work and with the latest one we're experimenting with techniques to cut that down some. Going well so far.
To go with the new body we're again working on the updated rollcage that works better with the new curved windshield. We'll be doing some rigidity tests next, including with side tubes removed to make entry easier. I'll report on the results when we have them.
10/06/15 While Tristan and Jay are cranking out Drakans and I am working on the Stiletto and a couple other things, Mark has been setting up a D2 experiment. The goal is to test torsional rigidity of the chassis, then cut away the upper main tubes for ease of entry, then test again to see what effect the change has on rigidity. We set up our original prototype chassis for the purpose. The front is clamped to the iron table (at 25,000 lbs it makes a very stable reference surface!). In the back, a tube is placed under the cross-member to serve as a pivot. Elevating the back also makes sure that only the very front of the lower main tubes is on the table.
On the left is a 3' bar to which a calibrated weight will be applied. On the right is a separate bar on which a dial indicator is set. It is important to have the bars separate, since the loaded one flexes and would distort reading at the other end. For the weight we used Mark and 'calibrated' him on the corner scales. Happens to be 182 lbs, applied at a 3' lever, so 546 lb-ft of torque to the chassis.
The resulting deflection is 0.070". While applying the load we saw that the clamps at the front have some movement so the clamp compliance was measured as well and came out to 0.010".
So, a net 0.060" deflection at 33.5" which translates into 0.10 degrees. With 546 lb-ft torque applied for this deflection it's approximately 5,500 lb-ft/degree stiffness (7,450 Nm/deg). Not as high as some of the claimed stiffness numbers out there but who knows how those are arrived at. The only ones I've seen measured in a similar manner are FSAE cars and those seem to be in the 500-1,000 lb-ft/degree range. To put it in perspective, if one rear wheel sees a sudden 4g load, such as hitting a pothole, and assuming infinite mass in the chassis, it will result in 4" of compressive travel and 7,200 lb-ft of torque on the chassis. Further assuming linear response, that would be 1.3 degrees of chassis twist so approximately 0.8" of additional travel would be due to chassis twist. Not a realistic scenario since the lightweight car would simply rise and transfer loads to other wheels, but it's an interesting worst-case analysis. This is an important thing to note, because it's not so much the absolute stiffness but the ratio of stiffness to mass that matters. This is why stiffness is sometimes analyzed in terms of natural frequency, which takes both into account, but it's not useful to the task at hand to go into that analysis here.
The goal here is a relative comparison rather than 'absolute' number. So, the next step is to cut out the tubes. There is a specific reason for the locations of the cuts, it will become apparent down the line so just take it as such for now.
Now for the really interesting bit. The removal of the tubes had no measurable effect on chassis stiffness. We checked it several times. What this tells me is that with the rollcage in place, the center of the car is much stiffer than other areas (my educated guess is that the tail section is contributing most of the flex). This would likely be different in a chassis without the cage, like the Drakan. There are a couple other experiments we mght try next but the main goal has been accomplished. The tubes can be removed without any negative effect on dynamics. On the other hand, the change in the ease of entry/exit is dramatic. This will impact what we do with the new D2 bodywork, and is also a bit of data that feeds into the upcoming D5 project. Learn something every day.
Speaking of Drakans, another one is ready to ship and one more about a week after that. When these are done we'll have a bit of a pause before the next batch of 5, during which time many other things will get done.
Site Sponsors and Links:
- CMS Lap Timer - Advanced Track Timer App With Video Integration
Tire Rack - Tires, Wheels (use link when buying wheels/tires to support this site)