Help with hesitation at high RPMs - fuel starve?

[KnifeySpoony]

Ok so issue with my car today. Recent build 420R with caterham roller barrels, stock tune. Trackday number 8 with the car, has performed flawlessly until today. In the third session at Laguna today, I let the fuel get a little too low, and got fuel starve (in a right hander, I believe). POPOPOPOP for about 1 sec, so I quickly dialed it down and brought it in. When I filled up, it took 4.6 gallons to fill it. Not sure if that is the expected level to risk fuel starve. The track ran out of 91octane so they only had 101, so I filled up with the 101 (had also topped up with the 101 with about 3 gallons when I got there in the AM). 

 

So the next session, car starts/idles without issue. I go out - warm up lap, short shifting around 6k RPM, no issues. Car is pulling hard and smooth. Second lap, start to bring it up to pace. When WOT, as the revs hit about 6k/6.5k, it starts to hesitate like it's missing. I immediately lift, but next straight same thing, so lift again. After a couple more times and realizing I wasn't imagining things, I pulled in. Looked over the car, everything seemed in place. Next session, same thing - instead of lifting when it occurred I stayed WOT and as the revs slowly climbed,  heard popping from the exhaust. Also, if I would accelerate at less than WOT, the car would pull smoothly all the way to redline.

 

SOOOO, I would guess there is some kind of restriction/limiting factor in fuel delivery? Is it possible getting fuel starve somehow damaged the fuel pump and now it isn't able to provide enough pressure to go WOT above 6k? What else should I be looking at?

 

Thanks if you made it this far!

 

[Croc]

 

Random thoughts as I start to think through this:

• An S3 tank is 8 gallons (36L).  Fuel pump is located drivers side.
• A right hander should not have made any difference as the fuel would have been all on drivers side where the pump is.
• change in octane is not going to make a difference on a standard Caterham map - they run rich at standard.  Higher octane would have cleaned any deposits lurking in the fuel system - unlikely given car age
• Fuel gauges in a Caterham are notoriously inaccurate but you topped it up with 4.6 gallons which implies 3.4 gallons were left in the tank.  That should have been sufficient for no issues as it is usually only below a third of a tank you start to see fuel cut off from fuel sloshing around in the unbaffled tank on track (depends on the corner).  However, if it is a long banked corner (e.g. NJMP Lightning) then you want more than half a tank on board.
• Its not clear in your later testing whether the throttle hesitation was happening on straights or corners?  I infer from reading it was on both?  If both thats useful to know as it takes corners out of the equation
• Your later testing was on full(ish) tank of fuel so that would seemingly rule out fuel surge in the tank
• You could get to redline without hesitation/popping by avoiding WOT - that tells me fuel flow is working and engine is getting sufficient feed from tank and pump.  How far off WOT were you when you did this?
• Just want to confirm - you hear popping while on throttle AND you feel hesitation from the engine in a cutting out sensation?   How perceptible is the hesitation?  Trying to gauge if it is 1 cyl or all of them.

 

From what I can discern so far, I think the areas to focus on are:

- TPS failing or needs a reset

- coil failing at WOT

- maybe the roller barrels have slipped out of actuation alignment relative to correct throttle position?  Unlikely...

 

[KnifeySpoony]

I will review my onboard and confirm which turn the initial starve occurred in. I just put in ZZRs so definitely increased lateral loads. 

 

Yes, the later testing the starve was happening on straights, with a completely full tank.

 

I don't understand why slowly achieving redline confirms good fuel flow. At part throttle, wouldn't fuel demands be reduced? I probably was at 50-75% throttle when it was able to climb past 6k. 

 

The popping only occurs if I stay WOT after the hesitation occurs, otherwise, it just feels like I'm hitting a rev limiter. I would guess that it's more than one cylinder based on how abrupt it is. In fact at first I thought I had some ECU fault where rev limiter was kicking in early. Would easimap tell me if there was misfire on 1 vs all cylinders?

 

I will check TPS

 

I check roller barrels - they are sync'd and functioning properly.

[sf4018]

I had a very similar thing happen at VIR a year or so ago, it felt like it was losing a cylinder for a few seconds on the straights, then would suddenly be fine again. After that happened, I coincidentally did the fuel cell project, which required putting in a new tank, an always on external fuel pump, fuel pressure regulator with return to tank and a new ECU. Basically like a 420 race spec solution. Never happened again. During the project I discovered from the SBD folks that the road going fuel system for the 420 was tuned to work 99% of the time but is basically flawed, the flaw revolving around when and how often pump energizes given that there's no fuel pressure sensor. Basically fudged to work.

I'm not saying this is what is happening to your car but it sure sounds similar. I can dig out and post the video of the issue from VIR if you'd like to compare.

[KnifeySpoony]

Another element re: fuel pump/regulator- when I did the roller barrel upgrade, it came with the uprated green injectors, which I installed. It didn't run (I made a post about it) but I learned that they made a change to the fuel system very recently such that the roller barrel tune can run with stock injectors. Something about an auxillary external pump: 

Maybe that could be the culprit?

 

 

 

[KnifeySpoony]

Checked my video. It was indeed a right hander that I first heard the popping that I thought was G-induced fuel starve. If this is indeed the wrong type of corner (ie fuel pickup is on the other side of the tank), than that could certainly be a red herring in all of this - maybe the original starvation episode just happened to be in a corner and was just the moment of failure of whatever failed.

[Croc]

17 hours ago, sf4018 said:

... the flaw revolving around when and how often pump energizes given that there's no fuel pressure sensor. 

 

Can you please clarify or expand on this?  

 

 

18 hours ago, KnifeySpoony said:

I don't understand why slowly achieving redline confirms good fuel flow. At part throttle, wouldn't fuel demands be reduced? I probably was at 50-75% throttle when it was able to climb past 6k. 

 

You don't get to redline unless the pump is pumping enough for the engine to run that fast.  What you may have is a problem with the flow regulation to allow a wide open throttle.  I thought the 420R had a flow sensor but Simon has hit on something important.  

 

18 hours ago, KnifeySpoony said:

Would easimap tell me if there was misfire on 1 vs all cylinders?

 

 

Yes it should but you would need someone in the passenger seat with the laptop to detect in real time operation unless you can record somehow.  

 

Your description is implying it is an all cylinders event.  Rev cutout is a very distinctive feel.  

[sf4018]

4 hours ago, Croc said:

Can you please clarify or expand on this?  

The internal fuel pump is controlled by a module, the module when powered up sends a 1+ second pulse to the pump to charge the fuel lines. When the engine is started the ECU sends an output to the pump module to command it to energize the pump. This ECU output is not on all the time but on demand. The problem for the ECU is there's no fuel pressure feedback, so it has to estimate fuel demand (no idea what the algorithm is) to maintain fuel pressure.

[MV8]

Not an expert or fan of newer "crap", BUT, Ford places the fp sensor on the hard line on the chassis. Only "Direct injected" engines with a cam driven secondary high pressure pump has a second sensor on the rail.

If there really is no fp sensor, I expect the feedback is based around a collective injector pulse width expectation for a given cts, hot o2, rpm, iac, maf, and tps. Typical of older equipment is a prime, then wait until rpm is near idle, oil pressure is above a minimal psi, or an airflow switch to run the pump.

I'd rather have an fpr return or carbs.

[fastg]

On most cars I have played with including my Honda system. The fuel pump runs for 2 seconds when you turn the ignition on. As soon as the engine starts the fuel pump is switched on, it runs until the motor is turned off. I had this with both a return and a non-return system. Any fuel delivered to the engine is controlled by the injectors. 

 

Graham 

[IamScotticus]

I realize the OP is FI, but...

Last time my X flow ran, I had a cut-off problem on short& quick right handers, not necessarily too fast, but over 20mph.  I am on 40s with Facet diaphragm pump, no regulation.  Standard Cat tank, no bag or foam.

I never resolved the issue before the car got put away.

 

Since then, I have learned for the refresh, I need to verify my floats and needle valves, as well as pressure.  I will be going back to mechanical pump.  Im the one who put the Facet on and I don't remember having the problem with the mechanical pump.  Formula Ford uses it successfully...

I might keep an electric on for a primer.

Edited June 4, 2022 by IamScotticus

[KnifeySpoony]

Ok, so I have done some more testing. I don't thinks its a fuel starve issue. Pretty sure it's a misfire/electrical issue. First I checked all the grounds, visual inspection of wiring. Everything looked good. I filled up with 91 - tank took about 5 gallons, so still a bit of leftover 101 in there (fwiw, Bruce Beachman initially thought maybe the issue was the car not agreeing with the 101 octane). Went and did some pulls - at first it seemed ok on the first one or two pulls, but then starting missing above 6k, so I went home.

 

I read a bunch about spark plugs. People on the UK forum strongly recommend the BR7EFS plug - colder plug/smaller gap, supposed to be better at high RPMs, at the expense of rough starting/idling. So I swapped plugs today. Started instantly and idled without issue. Drove it around a bit to warm it up (ran fine), then did some pulls. First couple pulls seemed fine, then it started to miss again, THEN it started to act totally differently. It started to feel like it had a persistent misfire on one cylinder (different than the all-cylinder-missing rev-limiter feel thing it was doing before). At all RPMs it was doing this, persistently, so I limped it home. And here I sit.

 

Any thoughts? It seems like it got worse after things warmed up/ran for a while. Could this be a voltage/battery issue? Coil issue?

[sltous]

Do you have any kind of data logger?

[KnifeySpoony]

i have an easimap cable - not sure what else I need to log data

[Croc]

The good news is that you are making progress.  The bad news is that you are still feeling your way through the problem.  

 

In the air, spark, fuel equation you seem to be zeroing in on spark as the problem - so thats encouraging.

 

I would go back to your original spark plugs.  The problem with spark plugs is they are octane sensitive and tune sensitive.  You don't know if you are replicating their exact experience.  They could be running higher octane than you, for example.

 

As someone who is currently chasing an electrical cause engine miss in tighter higher G corners, I really can sympathize (73 BMW CSL Group 4 for me).

 

You have two options to proceed here:

 

1) Hit and miss and hit again - test electrical items one by one.  If you take this approach then the coil would be my first stop to test first and secondary resistance by individual coil or just wear the expense and replace, keeping the old one as a spare if its shown that the coil is not an issue.     

 

2) Take a logging approach.  You can do Easimap - needs the cable and a laptop.  But does not record - only does real time - so you need a person in the passenger seat to be monitoring the laptop to see what happens at testing points or fix a camera on the laptop screen and record real time.  The other alternative which I have done successfully is use the AIM Solo DL to log data, overlay it on video (via RaceRender app) and then review.  I think @sf4018 did the same?  If the ECU is the cause of the problem then a logging approach will be best to diagnose.  

 

 

  

 

 

[KnifeySpoony]

Thanks - I'm thinking logging may be my best best to help diagnose. I will definitely go back to the prior plugs and see if the behavior reverts, or I keep getting the persistent missing. That will be informative as well.

[Croc]

One other thing - when you took your first set of plugs out did the tips show evidence of running rich or leaning out?  

 

I am expecting you should see them fairly clean but would like to confirm.  

 

[KnifeySpoony]

Here are the original plugs - they look ok to me

 

 

[KnifeySpoony]

Here is the new set. The one on the left is the rear cylinder (pulled a day after driving, otherwise I assume that is the wet/dead cylinder?)

Edited June 12, 2022 by KnifeySpoony

[KnifeySpoony]

Maybe the consistent miss on the rear cylinder with the BR7EFS plugs is a blessing for diagnosis? ie, maybe should put them back in then do some troubleshooting. ie switch plug position, then coil position etc? Will datalogging show cylinder misfires?