More Suck, Less Starve

It's a problem with both the MS3 and MS6 - fuel starve under various conditions. If you have a high powered drag car, you'll suck the basket dry. If you have lots of lateral grip, you'll starve the basket if you're running low on fuel. Either way, fuel starve is a problem for many of us.

Option 1) Run a surge tank/swap in a fuel cell. Fine solution for lots of people, and they aren't terribly expensive. If you don't live by a rulebook, this is a great solution.

Option 2) Holley Hydramat. That's what we'll be talking about here.


Check out this video if you don't know what it is: Link






My rule set doesn't allow surge tanks, so that's out. It does allow fuel cells, but they have to be located within 6" of the original fuel cell location and can't be much smaller than original. With the stock tank being low in the car and made of plastic, it's in a good location and is already very lightweight. We've been running it full for pretty much forever to prevent fuel starve - 16 gallons of fuel is heavy. I'm hoping we can start events with 3-4 gallons of fuel, which would lighten the car by ~70lbs. That's huge.

On to the actual work - Pull the fuel tank and figure out where the hydramats need to go.

MS6 fuel tank layout:


With the pump being at the forward position of the tank, and only on the driver's side, it's easy to see where the problem comes in. I'm going to focus on two areas, though, since it's a saddle tank. Acceleration and turning in either direction puts fuel in the rearmost corners:


The saddle tank drives a need to pick up fuel from either corner, so I'll be using two hydramats. You can bend these things around a bit, so I want them to wrap up onto the sides of the fuel tank. I decided to use two of the 15x15" hydramats, sort of like this:





Sort of like this:














Keeping it secured in that location is pretty easy. You can do this two ways, and I'll be doing it both ways. Holley makes little magnet adapters that sandwich the tank. These magnets are incredible strong:



Hydramats have attachment holes on the corners:



What that looks like on the inside and outside:



I will also be adding foam into the tank for two reasons - 1) securing the hydramat and making sure it stays pushed into the rearmost corner. 2) Help prevent fuel slosh. Big box of foam:





I'll talk about the fuel pump and plumbing situation in the next post.

 

Nice! I bet these pads will make a huge difference.

Though, never seen foam added to a tank before. Is that specifically made to be used with fuel?

 

Yeah, the foam is compatible with gas, ethanol, and methanol. I picked them up from Speedway Motors.

 

so these are basically a surge tank you can put in the gas tank?

 

That's not its main functionality, but it does pull from its internal reservoir before it will suck air. The real benefit is that if any part of it is touching fuel, it will pull it in without sucking air from the exposed areas. I think this video shows it better than the first one:

 

That's freaking cool! Video definitely helps

 

Alright, so plumbing this should be easy, amiright? Like most things, it's not.

Here was my original concept.



The Holley ITFP adapter (16-137) should have been the easy button. I would love to keep the ITFP because it simplifies the plumbing. Issues I have right now:

1) The Holley adapter doesn't work with OEM, Deatschwerks (DW300C), or AEM pumps (50-1200). Seemingly, it doesn't work with any typical 255lph ITFP that isn't Holley brand - and the Holley brand isn't compatible with E85. The adapter doesn't allow the pump to seat against its O-Ring for the fuel pump inlet - the raised channels/nubs prevent it from seating all the way down.

I prefer using OTS parts, because it simplifies quick fixes that we sometimes do at events. I could modify the pump by removing the nubs. I could also modify the adapter so it accepts the nubs/channels. TBD on that front.






2) Pressure regulator. The OEM setup has the pump in a plastic sleeve that has a regulator attached to it. I can't use that sleeve anymore because of the Holley adapter. This is going to require an external regulator that I haven't decided on or figured out how to plumb yet.



More to come on this front.

 

I think we have a winner - Walbro GSS342G3 if anyone wants to venture into this. I'll submerge this and test to make sure it doesn't pull air past the O-ring.

If the other fittings get here this weekend, I'll plumb it and do some testing.

 

We did some testing with the hydramats over the weekend to get a feel for how they worked and if they had any weird issues we need to be aware of. We mostly did this with the mats in a couple large bins, and we just started to test it in the tank.

1) The hydramat pickup will pull air pretty early if the pickup angle isn't perpendicular to the mounting surface. I think this will be solved by supporting the area with the Speedway foam.

2) Early testing with the mats in the Mazda tank show it sucking air from what would be the inside half under turning conditions. More to come on this, we need to work with mat placement more before we can conclude anything.

3) Most surprising to me - the fuel pump can suck some air, but it doesn't transmit to the downstream fuel line unless it's sucking a BUNCH of air. Check out the 2nd video.


Showing how sensitive the mats are to pickup angle:





Running 1 dry, and sucking some air without pushing it downstream





Angle testing 1 mat to see how it picks up from a single corner touching fuel:

 

The fact that it can pick up fuel from just that one single corner of contact, is pretty amazing!

 

Things are getting...complicated.

I can't trust that the hydramats won't allow air through in dynamic situations. They suck air during transient tests by lifting up one side of the tank and sloshing the fuel around when there's low fuel. Transient maneuvers is about 100% of autocross, so that won't work. Let's complicate this shit.

Before, I had a lift pump on the passenger side into the stock basket before, so I'm doing a more robust version of that now combining it with the hydramats. Effectively, I'm making a larger basket that the hydramat system will pump into, then pull from the basket with a second pump:



And I just got the materials in for the basket:






I'm nearly finished sorting out what hardware is required to make this all work, and I'll post up some details once it's all set.

 

Progress. This is the final conceptual drawing (hopefully):



I've been busy making the new basket. Disregard the rest of this post if you don't care about the fabrication, but I like to make stuff so I'm going to show how I did it (lol).





I just started with the pics from my post above - 6061 tube and flat plate.

Turn the OD and ID so it fits into the stock fuel tank opening and you maximize internal volume.



Get a rough idea of how you'll cap the thing with the flat plate. I chose a bolt pattern of 12 screws, with a 13th so it can be aligned the same way every time.



I opted to start with a thinner tube than what I could have, because I'd rather weld on some little pads and machine them, instead of having to turn the ID down a whole bunch. In the end, probably didn't save a ton of time, YMMV. Weld blobs:



Machine those bad boys down -




You could drill all the holes in the cylinder, and then the top plate and hope they line up perfectly. Or the method I chose - tack weld the top plate on, drill the holes through both to ensure alignment. I tapped through both the plate and cylinder at the same time, then opened up the plate holes to have clearance for the threads.






You probably noticed that the top plate isn't round. I chose to turn the top plate down after having the attachment complete so I could ensure concentricity:




The top plate needed a bunch of work. There's 3 fittings, and I chose to put a level sensor in it for future datalogging purposes, and just a gauge for now. Not a lot of pics here, but here's the top plate with a bunch of holes in it:




The level sensor is just a simple reed switch type.



The 6" version just barely clears the bottom of the basket:




Anyway, I wanted to be able to pull the level sensor if needed. That allows a view port if I want to see what's happening in the basket. So that requires some more machining...

Make square things round:



Transfer the fuel sender bolt pattern and drill holes (SAE 5 bolt patterns are not equidistant...)



Tap:



Test:





Attach your donut to the bottom of the plate. Tack in place:




I cut the donut hole undersize so I can make the top plate hole and donut hole concentric (seems to be a theme). Swap over to the 4-jaw chuck and cut both together:



Top plate is done:






Time for the bottom plate. Similar idea to the top plate, but the bottom will be fully welded:



Turn it:



I turned that on a bevel, so not unexpectedly I cut into the seam and had to reweld:



And turn again:



Now it holds liquids


Quick work to make a weep hole in the basket so it spills back into the main tank. I put that on the front side of the basket. There's pros and cons to putting it anywhere, but I figured I wanted as much fuel as possible during acceleration.





Assembly. From here, we just attach the bulkhead fittings and fuel level sender:





And drop it into the tank:

 

Ran into a couple complications. When I put the tank in, it raised the floor of it a bit because of the strap running underneath that area. I was ~.150" too tall at that point so I needed to trim the basket a bit. I turned .050" off of both the bottom and top, but that didn't quite get me there. So...

Cut in half. Since the parting tool is .118" wide, I didn't turn any additional from the middle.





Weld back together:



^^I turned that flat so it still fit into the tank.


I also modified the passenger side hat so I could pump fuel out, and return fuel to it. You have to remove some ribbing on the underside to get a flat surface.




This is what both sides look like with the tank back in the car and all fittings attached (sans fuel level sensor on the driver's side).




We started testing a bit last night but haven't fully vetted the system. Here's the internal pump filling the basket, though...



Don't think we'll ever have problems draining this thing, lol. I am a little concerned about just how fast it's filling, so I will likely open up the bleed hole at the top of the basket so it doesn't pressurize it.

 

Do you have an idea of how much fuel you burn with each run and how that relate to the volume of this new tank?

 

The overall volume of the tank is unchanged, I can still put 15 gallons into it if I wanted.

But, I think we burn about a gallon per run, and the new basket is right around 0.5 gallons. With the constant fill from the ITFP and return, it's never going dry lol.

Once we get the car back on the ground, I'll pump it all out and do a total fill to get an idea of just how empty it really gets. [Also need to weigh the car empty because I'm really curious how much weight we've dropped this year.]

 

I didn't think overall had changed much was just curious how important the fill rate on the new basket would be and it seems like not that important

 

For sure, a 255 pumping at effectively 0psi is wayyyyyy overkill. Not something I had considered beforehand, and it has me rethinking how the return line routes. More to come on that as we test things out.

 

Some minor updates, but most importantly - the car is RUNNING with this new setup.

After thinking about the ITFP being in free flow mode to fill the basket, I decided that we would try it out with the elimination of the return line pumping into the basket. I already had everything plumbed, so I just capped that section and I can always go back.





An updated drawing of where we stand right now:




We did all tuning on this setup 2 weekends ago, but a stationary car isn't a good test for what we're trying to fix. [It worked fine, btw.]

We took the car out for its first event with this setup this past weekend and it performed well. I do have the basket level sensor and gauge hooked up, so we could watch the level during the runs (and it's in view of the camera). At worst, the gauge went 1 notch down from full during some quick transitional elements, but no fuel starve!



I did find the outside surface of the basket covers (what you see under the rear seat panels) were slightly damp after a run. It seems that the PTFE washers aren't fully sealing, so I have a couple options headed my way to fix that. Outside of that, no issues!

 

Coming back to this since it's fresh in my mind from the weekend. Before the event, I pumped out all the fuel so we could start with a known amount. Then after the event we pumped it back out to measure how much we burned.

Start - 10 gallons
End - 7.9 gallons
# of runs - 7 (all about 40 seconds, so a short course)

So we're burning about 0.3 gallons per run on a 40 second course. It should be roughly proportional for longer courses. We have a 3 day event this weekend (tnt on Friday) so we'll do the same exercise starting with less fuel and running it way down.




Going back to the decision to omit the return line from feeding into the basket. When I pumped the fuel out of the tank for the first time, we pumped through the entire system by hooking up to the hard line at the firewall and pumping into a jug. Essentially the external fuel pump was in free flow mode so we could see how the 2 pumps interacted with the basket level. The ITFP is able to out flow the external pump and keep the basket full, so this setup is definitely good for the K04, and should be good with much higher power levels as well.

 

Developed some issues over the weekend trying to run it lower than before. We had only gotten it down to 8 gallons last weekend, so I started with 5 gallons this past Friday. That almost immediately had issues, only 1 or 2 runs in. We heard the pump starting to whine while idling in grid (thankfully not during a run). Friday was a TNT, so we wanted to maximize our time on course and left diag for later - so we added 3 gallons to bring us up to a known good point and it resolved the issue until ~9 runs later.

Rough math says that it's starving around the 6 gallon point. We just ran it pretty full to get by. After we ran the car on Saturday (30 runs!) I topped off the car and had a bulkhead fitting on the passenger side start to leak when we were working on the car in paddock. Since it's at the top of the tank, I think what's happening is that we have a net flow effect that's pushing all fuel to the passenger side, and the driver's side well is going completely dry...

So a little bit of diag is needed. I'm going to send my endoscope into the tank to see if the weep holes in the basket are doing their job. We might need to open them up, or we might need to reconnect the return line to the basket. If it's not that simple, I have some ideas but I want to start with the simplest first. We have a couple weeks until the next local event where we can try this out, and our first big event at the end of July where I really want it to be solid.