K24Z3 endurance build - Vegas
 

New thread to document It's conversion from turbo BP and AZ6 for S1 to K24Z3 & Quaife sequential for endurance racing. Focus will be on AER, Lucky Dog Racing. Possibly some NASA WERC, Champcar EC class and maybe WRL if we travel east with it.
Vegas was originally built for Supermiata S1 class, finished late 2018. Raced a bit early 2019. Won Super Lap Battle Unlimited RWD in 2018 with it (boost turned way up). S1 (220whp turbo) class never really took off the way S2 (140whp N/A). We tried to future proof
it by including connections for lights everywhere, radios and such in the scratch built harness. Cage is sorta like an FIA cage, far more robust than a traditional Spec Miata style floor mounted 6pt cage.

For enduro, we knew we could improve on the turbo BP and AZ6 drive train. As solid as the turbo kit proved, to total system was still mot as bombproof as we would want for enduros. The AZ6 also a liability in that regard. Beyond that, fuel consumption was the Achilles heel of that drivetrain config. In S1 trim, it was too fast for pretty much any budget enduro series so we didn't need all the power it was making. So we decided to switch from E85 to pump gas, fit a 11.0:1 K24Z3 motor and bolt it to the Quaife, This should drop us from about 16 gallons/hr in S1 trim to about 10 gal/hr and be more reliable to boot. We figure 180whp at around 2200 comp weight should put us in the right window for enduros. One key strategic goal is to be able to run the maximum 2hr stint length on one tank. We'll be slightly short of that, but can probably stretch it out by short shifting if we find ourselves with a gap to 2nd place late in a race.

We're using the Kmiata kit for the K24Z3 swap. The Quaife needs a low MOI clutch/flywheel assembly to allow two pedal driving. By two pedal, we mean flat upshifts and clutchless, blipped downshifts. To downshift, you left foot brake and tap throttle to unload dogs while shoving lever forward. Clunk! Like butta. Being able to shift up or down, anywhere at anytime, instantly and without upsetting the car or worrying about a mis-shift makes driving smoothly so much easier. We decided to design and make our own proprietary flywheel to mate the K24, Kmiata adapter and Quaife our existing 7.25" Racetwin clutch.

We plan to run the exhaust down the pax side, deleting the PPF. We'll make our own trans & diff supports. The PPF becomes a failure point in endurance racing, hanging low and vulnerable. Only needing a 2.5" exhaust at our power level makes routing a bunch easier. Aero will be pretty much the same as S1.
We should be able to run a smaller oil cooler. The huge Setrab we have now is overkill for a K24 on gas running stock redline. By raising the vent tube in the gas tank, we can squeeze a full 15 gallons in their including the filler neck. We have been doing that since 2010.

https://cimg7.ibsrv.net/gimg/www.mia...f311b8c79f.jpg

K24 7.25" Racetwin flywheel & clutch
 

https://www.miataturbo.net/suspensio...-trans-102233/

The flywheel we're making to be able to use our existing 7.25" Racetwin clutch.

https://cimg0.ibsrv.net/gimg/www.mia...6a1d2772d8.jpg
https://cimg8.ibsrv.net/gimg/www.mia...cc0b3aadd0.jpg
https://cimg9.ibsrv.net/gimg/www.mia...40256f310c.jpg

If the quaife bolts in place of an AZ6, does that mean this flywheel could be used with a k24z3 and an AZ6... :fael:

Looking forward to seeing this come together!

Yes, of course.
If if you are interested in one when we do the production run, send an email to info@949racing.com and ask to be put on the list. The only plan to make 10, five of which are already spoken for.

Are you guys going to be doing the same kind of oil pan/ oiling system modifications as on Deviate?

I think the PPF only becomes a failure point when you cut a very large notch out of it :giggle:

Excited to see this all come together though, especially the clutch/flywheel, good shit there.

No need. We're not revving it nearly as high. V-Tec requires a lot of oil volume, no so much pressure. OEM K oil pump cavitates badly when spun well past OEM redline. Long oil tube also has poor .cf. Keep revs lower and it's not much an issue.
PPF bolts and frame hang low and get hit when/if the car goes off track. Particularly vulnerable when car drops two and high centers on track edge. We have seen PPF failures in sprint and enduro. Aside from the lower flange on a Racing Beat header, the F&R PPF bolt heads are the lowest points on the chassis. Ditching the PPF make service easier as well as making exhaust routing far easier for a K.

While I understand the desire to eliminate the PPF because of the transmission and other plans, but the Mazda solved the issue with the bolt heads (before removing it as a cost reduction) with just an aluminum wedge forward of the bolt heads. Been using one of those PPFs on an enduro car for years with no bolt head damage.

A bit more detail on our plan and overall race strategy. We wanted to build a car that could run at the front of the fastest class but still go 2hrs on a tank. Depending on the series (fuel tank rules), this is basically impossible with the OEM tank. No way to get down to the 12-13 lbs/hp or so we will need and only burn 7.5 gal/hr. Most of the faster cars wining Champcar and LDR are fast but not running 2hrs. We won't be able to either so at least its a level playing field. I think LDR allows an aftermarket fuel tank so the possibility remains of dropping a 24 gallon fuel cell in the pax floor. For Champcar we're in EC no matter what so we might as well. We'll get it running with the stock tank first though. Then we'll asses our competitiveness and see if a full cell is warranted.

Still the lowest part of the car by far. The load from any impact goes into the PPF. I can see both side of the rationale, I'm choosing to ditch it.

Isn't the goal of 180HP a little low considering the 09 TSX the Z3 comes from is listed at 201HP? What is your intended engine management? Are you planning on keeping the NB Diff (assuming you have an OSG in there) or move to a Getrag or RX7 unit?

Very interested in how this turns out, and i'm sure there will be a lot to learn from this thread.

180whp vs 201bhp. focus is efficiency. Won't be surprised if it comes in over 180whp though.

Just doing some though exercises with the part of the team that has remained active in enduros the last few years. Don't fixate on power for this project, we're not. We have the luxury of tuning up or down as needed. As you know, there is such a thing as being to fast in budget enduro, depending on the series.
Goal is to have BSFC as low as possible. So free flow intake and exhaust, pump gas, perfect oil and clt temps and a careful tune. On paper at least, we should be able to match the gal/hr of the best BP engine Miatas but have another 50whp or so.

As mentioned earlier, we'll see how the package works in race situation. If its too slow and the fuel economy isn't making up the difference, we'll turn up the wick. I think it'll make about 210whp on gas as is. Cams would bring that to 225 or so. But we hope to be competitive with a junkyard motor, revving it to 7400 for best BSFC. We'll see.

Haltech 2500 ECU. DBW, sequential shift stuff, Vtec (yo)
OEM diff with Supermiata tuned 3.9 OSG. Utterly bomb proof at 400whp with the cooler so no concerns in the 200whp range.

Our test car make 180whp on the first pull with a 2004 TSX basemap, and in 30 mins of tuning we had it making 206whp. There should be some more in it with a bit of intake length tuning, plus our engine is a $300 junkyard block with over 150k on it. I'd expect a fresher engine to do a bit better.

Yup. We published what we think we can run at strategically, not the peak it will be capable of. I'm sure cams, valve springs, E85 and leaning on it would net 230whp or so, but that's not an rock solid, high economy enduro config.

The discussions on the private Team FB group are to have a long list of maps to choose from in the lap top that travels with the car. From about 160whp with 6800rpm redline to Kill, in 15whp increments :D

Please tell me where you've found cams for this motor. Bisimoto had some on the website but they're gone now...

I was more concerned about the diff ears, but if it hasn't snapped in Vegas after the buttonwillow jump, I guess it should be fine.

This gives me high hopes for my 60k mile motor...

I haven't even looked at cams because I doubt we will ever explore that area. Healthy junkyard motor will meet or power goals. My point was just that we are aware that more power is easily available if we need it.
It's scrambling everyones thoughts that our power goals are so modest. BUT ALL THE POWERS!!

https://supermiata.com/Kmiata-differential-gussets.aspx

Yes LDR allows fuel cells, nothing over...22? 24? Just to keep people from making 500hp and running full 2hrs by just running giant tanks. Not the most effective or restrictive cap but it is what it is. We ran close to 200hp and a few tweaks to the filler with great success, there were times we wanted more, but a few cautions would keep us on schedule. Green flag for 2 hours straight is what we couldn’t do, and why we decided to put a cell in the trunk. I think I already rambled about this on FB though, but im bored on the dyno.

Were you monitoring EGT? I'd be interest to see what lambda/equivalence ratio, ignition timing, and EGT you're running. I'd run the highest RON fuel available with the lowest ethanol %, which will provide the best balance of energy density and knock toughness. OEM fuel enrichment on newer North American market cars requires running stoich, unless you need to enrich for component protection. Euro6D is already requiring RDE (real driving emissions), so in order to pass CO, they're basically operating at lambda 1.0 continuously. Also it's quite common in racing to have different fuel enrichment maps- economy to full power. The Haltech 2500 allows for a second map, so you would at least have the option to change strategy mid race.

Enrichment will have a yuge impact on fuel consumption. Control fuel and IG to run ~900C at cylinder head outlet by leaning her our and relying on your knock control system, as much as you dare, to run close to knock. This is one area where the OEM engine controls far outpaces the aftermarket (ignition control, knock detection accuracy, exhaust temp prediction). IG will have the greatest impact on exhaust temp. It's an iterative process: Lean out 0.05 lambda, advance to knock, check temp, repeat. I wouldn't run a stock K above 930C sustained, at least for endurance. I also wouldn't even bother trying this unless you can run steady state on a dyno and map each gridpoint.

One thing that is coming out of RDE is DBW throttling. Since they're running stoic, or very minimal enrichment, at peak power, the DBW is being closed at high engine speed to control exhaust temps. Also keep in mind that engine speed is not a friend of fuel economy - downspeed when possible. This has as much to do with the driver as it does with engine architecture and calibration. Audi's LMP1 was famous for extreme fuel economy measures even as far as banning left foot braking to reduce time on throttle and also coasting into the brake zone. Not useful for qualifying, but the driver plays a big part in endurance racing to turn the available fuel into the most efficient lap times.

Yup. Something we learned in the T25 days was that is was far more efficient to stay WOT and just keep shifting just past the torque peak. Thats where BMEP and BSFC will be highest, more or less. Lower pumping losses. using high revs and "feathering the throttle" sucks fuel. Thus the plan to run it at low revs.
My guess is we'll be shifting somewhere between 5200-6200 most of the time.

Josh,
I recall the same reg on fuel cells in LDR. There is an upper limit IOW.

Goal for now is to just get it running on the stock tank, learn how to tune it. Get the team used to the car, see if its competitive or not. If not, we reassess and see what steps we need to take to make it competitive.

Following with interest :)

Camshafts & Gears | Heads | 4 Piston Racing

Those are designed for the A2, not Z3. Z3 has a wildly different head setup.

Ah, interesting. Good to know.

Streaming live data
 

Just learned about this. I think we'll add this to the build. So many benefits from having this data on the pit wall.

https://www.autosportlabs.com/produc...ems-to-podium/

This is gonna be a fun combination! It’s pretty much the ideal drivetrain. I’d take a lower powered engine with Quaife sequential over a high powered manual anyday. It’s just matches the fluidity of the Miata better.

Glad to hear Vegas is getting repurposed. That shell is well prepped (stripped, seam welded, FIA cage etc) Great strategy to maximize fuel consumption. Does the ECU/VTEC have enough control to run Atkinson cycle during maintenance throttle? That would really help stretch the tank during cautions etc.

Once again many of us get to live vicariously through you.

Dunno on mapping. From our experience with N/A BP's, we know we can get away with really lean mixtures at part throttle/light load. The K is so much more resistant to det.

I think the combo of DBW and the sequential will make it supremely easy to drive and place on the track. We have an EPS unit here but I don't think it'll fit without some significant cage revisions. So we'll just dial the caster down.

Have you considered an electro-hydraulic pump like an MR2 or Mercedes unit?

Nope. Not interested in embarking on a new R&D project. Just doing our right side exhaust and the flywheel are keeping us busy.

Finally getting to this stage. Project was on hold for a while. Dropping long block & Quaife in so Renderos Racing can fab the exhaust and mounts. We're deleting the PPF and routing exhaust down pax side. Spoke with the guys at Haltech and settled on the 2500 series ECU. At least one US tuner has mated a Haltech and a Z3 so we'll have a starting point for a base map.


https://cimg8.ibsrv.net/gimg/www.mia...f9e4e3dcfc.jpg

Huh, not sure how I missed this build, but very relevant to my interests.

I found their data capture software and UI to be ... shite. I guess the hardware was fairly OK, but I found the software incredibly frustrating to use. I tried it on iPhone, an Android tablet, and my Mac. They use a UI abstraction layer instead of writing code for each platform and it seems to end up on screens or in dialog boxes where it doesn't know how to behave. I would approach with caution. I'm done with my RaceCapture device, I'll send it to you if you'd like.

This is good to know. They explicitly state on their website that they support the Z3. I haven't been able to establish the same for MegaSquirt. So far I've found "yeah, it should work", but not "I'm running it and it works great".

Without spoiling any secret sauce, what do you see as the strongest motivation to go DBW? I haven't decided on a throttle for my build yet, but may go the same route because someone smarter than me has already done the thinking.

Something to consider on the DBW setup for ~180rwhp. The WF guys ran into issues where the throttle blade will be open at say 35% to hit the target power for that RPM, the vacuum within the manifold can actually pull the blade closer down to say 25%. Their quick fix was to short-shift as it only seemed to happen at high RPM. So in this build's case, it probably won't be a problem. But hey, might save some time later if it does happen?

They talked about it on this Slip Angle podcast: https://www.stitcher.com/podcast/slipangle/e/75408329

That's good info. They're running K24A2's though right? A little harder to choke those ones down. I've got the Bosch electronic throttle body here that Kmiata offers...

Double Post, thanks IB.

They have at least two cars running A2s (Leichty and Manocherhri) and at least one car running a Z3 (Tabb).

DBW for both power management and to complement two pedal driving with the Quaife sequential. We don't currently have plans for flappy paddles, just going to use the load cell built into the shift lever.

4 Piston Racing has pre-orders going for their CNC heads for the Z3...

Any reason for the 2500 over the 1500? As far as I am aware the only differences are 4 extra ignition and injector outputs, along with an additional knock input. Those seem like they are not needed on a K-swap.

I have been happy with my 1500 so far. Definitely has some quirks and there are some things I think the MS does better.

Nope. But now we're investigating the new Nexus R5 which has integrated power management and full data logging. The data logging would be redundant as we'll use the Aim MXG for that. But the power management would allow us to delete the Racepak Smartwire. We're having to build pretty much an entirley new harness for the K24 anyway so de-pinning the Smartwire for the Nexus swap isn't too much extra work. John's looking at the available I/O differences to see if we'll be able to run everything I'd like to run.
We don't yet know if the Nexus can com with the MXG through a single CAN. Hope so.

From memory so I'm probably missing something

Power management
engine main power
fuel pump
diff oil pump
trans oil pump
fans for both diff & trans coolers
radiator fan
radio
driver suit refrigeration system
headlights, LED, FTP (flash to pass)
tablet for driver info on center console
wipers, 2 speed
heater blower
under hood light
interior light
inside trunk light
running lights
brake lights, triggered of pressure switch
Possible EPAS if we can figure out a way for the GM motor to fit under the dash bar
Gear position sensor
Telemetry transmitter

ECU functions /logging
Bosch temp/pressure combined sensor for engine oil
Bosch temp/pressure combined sensor for engine coolant
brake line pressure, single
3 axis accelerometer
GPS speed
wheel speed from LR ABS tone ring
Diff temp
Trans temp
Gear position
Sequential ignition
Sequential injector
Cam angle sensor
Crank angle sensor
Knock sensor

We're starting to add fuel pressure to most builds as well, recently installed a Link G4X on an E46 and can was easily setup between the MXL2 and the Link. However with KO Racing in my back yard, who has some great experience with the Link, it was an obvious choice for that build, my guess is there's a local with Haltech knowledge in your area. With ECU logging and AiM logging, I try to keep the driving information on the AiM, and engine information on the ECU, shared only when necessary, like TPS for video overlay.

If the Nexus R5 was out when I bought my 1500, I may have gone with it instead.

I hear on the grapevine from teh aussie circles I frequent that the nexus line will be supplemented soon with more varieties that may include less PDM, and fewer drivers etc for the more budget driven/road car driven consumer, in the same way the elite line was rolled out, with the flagship 2500 first then the stripped down versions.

Fuel pressure on the list already. I like to log it for diagnostics. Previous builds have logged fuel pressure. List was from memory so I'm sure I forgot several channels and functions, including TPS target and actual.

That would be nice. The R5 is appealing, but even my penchant for over-engineering doesn't extend my K-swap budget out that far...

Alex Renderos, Renderos Racing has changed his business model. Moving away from custom fab work and doing more serial production of his own line of parts. Vegas will be his last customer car. That said, it's taken a long time for him to get caught up and start on it. At least now,
it's the primary project. The other delay has been the special K24 to Miata trans flywheel for our 7.25" twin disc clutch. We have gone through no less than 4 different suppliers. All of which said "no problem", took the order then months later said "just kidding", we can't make it. Argh.
Still no solution on that end. Plan B is just running the standard (heavy) Kpower flywheel for a BP clutch with the Quaife. We'll see.

Anyway, Alex just posted some pics of the custom motor and trans mounts in his 'gram. We're lowering the engine and moving it back relative to the standard Kpower location. Also running exhaust down the right side and eliminating the PPF. So may custom mounts are in order.
Progress!

https://cimg7.ibsrv.net/gimg/www.mia...2fc285d641.jpghttps://cimg9.ibsrv.net/gimg/www.mia...2f5398dce7.jpghttps://cimg4.ibsrv.net/gimg/www.mia...c1475ba219.jpg


https://cimg2.ibsrv.net/gimg/www.mia...c88d9187c6.jpg
https://cimg7.ibsrv.net/gimg/www.mia...509625b30f.jpg

Good to see progress here; despite your setbacks, you may be on the track before I am.

I wonder if you'll have any better luck getting the flywheel done as a one-off (or two-off) than as a production run. It seems like Silicon Valley is full of little one-man shops like the one I sent to Ed; I don't know what he's capable in terms of production runs or certifying to standards, but he does one-offs and piece work for our company. Maybe at 10 or whatever someone his size wouldn't touch the project, but at 1 or 2 maybe you could interest them ($).

I'm intrigued by the idea of moving the engine down and back, but won't that conflict with the Kpower oiling setup?

No plans on serial production of the flywheel. That ship has sailed. Even getting one made has been tough. In the end, it'll just be a really expensive part, like 3x what we paid for the f'ing long block. le sigh..

So far, according to our initial measurements, we have a little room to move the motor back. Down, not so much. We'll see what Alex comes up with.

Yes, sorry, I wasn't trying to bring that dead horse back. What I meant is that I would have thought getting one would be straightforward albeit expensive. Sometimes my words/fingers connection is intermittent.

This was a fun notification to get.

You should try PrusaSlic3r instead of Cura. Pathing is a lot better, smarter settings choices instead of the nebulous options of Cura, I find the workflow a lot faster as well based on how they have the presets.

Unless you have an ultimaker... But thats another conversation lol.

As I tend to do with these types of projects, I've aimed a little higher in terms of performance.. and added a bunch of complexity to what should be an otherwise simple project. In this case, I want to move the engine behind the steering rack. Beyond the weight distribution gains, the engine will fit under the hood better and allow a simpler, higher capacity oil sump. The section of the subframe that's cut out is right under the oil pump. That front bit of the Kpower sump hangs lower than it needs to. That front drop section is just for extra oil capacity but since the pick up is in the rear section, that oil capacity is superfluous. So we'll cut the sump shallower there so it's just tucked under the K20 oil pump. That will allow us to tie the two halves of the subframe together. You can see that Alex has fabbed a simple plate to temporarily keep the rack in OEM position while he's working out the rest. That structure will change when it's finished. I don't know exactly how far back it's being moved, l left that up to Alex. I just asked if he could get everthing behind the steering rack. Looks like 6-7" maybe.

We're also keeping the exhaust on the pax side and ditching the PPF. It'll need a shorter custom driveshaft.

This tub isn't Vegas. It's a scrap tub Renderos has laying around to chop up and experiment with.

https://cimg4.ibsrv.net/gimg/www.mia...35228858bb.jpg

It looks like the firewall has been clearanced for the intake manifold.

https://cimg6.ibsrv.net/gimg/www.mia...d007c8a235.jpg

Well that's an interesting engine position! A few install compromises but gain several performance traits like you mentioned. Any class rules about fire-walls that would be a concern with this build?

One tip because I recently ran into it myself. On the intake manifold you have there, the seat where the nuts or bolt flange isn't machined flat/smooth. This will cause the stud or bolt to snap off on the IM and give you a fun vacuum leak(I was chasing mine to finally find out it is the IM). A common resolution on this(from WF) is to flatten the seats at the nut/bolt-head by hand, then use a nord-lock washer to help with any remaining misalignment.

I wondered about the utility of the pickup-less front sump. I asked David about this directly and his answer was that oil from the front was constantly spilling into the rear. Interesting to hear another take.

This is definitely going to be an interesting build. I mean, who doesn't love taking an engine that barely fits into the engine bay and moving it so it doesn't fit into the engine bay at all? :giggle:

Certainly no harm in having that little front sump on the Kpower kit but it's of limited utility. Unfortunately, it's just a huge 4cyl engine. 35% more displacement, much steeper ports make a much taller head. I may come to regret the decision to relocate it, we'll see.

WRT to rules, we can't even run EC in Champcar with the Quaife. We're mainly focused on LDRL (Lucky Dog Racing League), which allows the sequential. Maybe a few NASA WERC but we'll get slaughtered there. Weeping because we don't have WRL in Cali (yet). Depending on how well things go, we have dreams of shipping the car to friends in Columbus, OH or Atlanta for a few months so we can do some WRL events and maybe a GLTC.

The firewall cut and sequential are both a no-go for GLTC.

In GLTC they'd probably let you run as exhibitionist for the firewall cut, but no-go on the sequential. It's been a topic of discussion for a while, dog-box and PDK/DCT/etc are also being discusses(they're currently allowed without a modifier).

I'm really surprised WRL isn't in Cali...... too many other series out there maybe? Also, I've been meaning to ask Steven about convincing you to come out to Miata Reunion next year at Barber..... and possibly Miatas at the Gap the next weekend! :party:

Barber at the end of July? That's peak swamp ass season.

Change of plans. Decided not to go forward with the engine relocation. Car has been in the shop for two years already. Too much complication, design work and unknowables with the relocation. We just need to get it done. So we're just going to install the standard Z3 kit which by now, is well proven. Lots of WRL and GLTC builds running reliably in competition.

We have so many projects planned using limited resources that it's time to cut bait. Come hell or high water, I want to have Vegas tested, dialed in and ready to do 12hr enduros for the beginning of the 2022 season.

Many projects fail because they fail to reach MVP. Good refocus.

I could learn something from this. But I probably won't.

Was about to jump in here and suggest against the extra work. LDRL has put a serious cap on lap times, no need to make a car any faster than what a 200hp Miata with good suspension and brakes can do with a decent driver.