Any metallurgists out there?

[Wayne M.]

I was just talking to a coworker who did a conversion, a inline pump 12 valve into a 2 gen. He monitors low side fuel pressure, and said he gets extreme spikes that pegs the gauge and he said even broke a gauge. He's trying to diagnose the problem, but when I heard this, I immediately thought of this thread. Could a pressure spike be blowing the filter apart?

 

[GAmes]

It isn't the filter that is failing. The spikes are normal with a manual lift pump. Advise your friend that a proven method is a needle valve, barely opened, between the port and the gauge (or sender). On mine I drilled and tapped the banjo fitting and attached a flex hose from a grease gun. At the end of the hose is the needle valve and the sender for my EV2 gauge is attached to the other side.

 

[Wayne M.]

Spikes that high?

 

[GAmes]

Spikes that high?

The spikes are violent enough to ruin both mechanical and electric gauges.

 

[The patriot]

I have mechanical gauges hooked to my fuel filter (sorry...pics are long lost in one of the board upgrades)The spikes are pretty damn wild, but I couldnt imagine it playing a part in this.

 

[Joseph Donnelly]

Make sure it is not M10 x 1.5 thread, but if it has 6 marks it is grade 8 American thread. Metrics will have a grade like 8.8, 10.9, etc. on top. You can upgrade to grade 9 if you want:
http://www.mcmaster.com/#standard-cap-screws/=13vobsf
I doubt you are overtorquing the bolt; the grade 8s hold 35-40 ft lb easily.

 

[brods]

I was just talking to a coworker who did a conversion, a inline pump 12 valve into a 2 gen. He monitors low side fuel pressure, and said he gets extreme spikes that pegs the gauge and he said even broke a gauge. He's trying to diagnose the problem, but when I heard this, I immediately thought of this thread. Could a pressure spike be blowing the filter apart?

You might be on to something. How high do the pressure spikes go and what is the diameter of the open end of the filter housing? Once the force on the bolt from the spikes is calculated then the frequency needs to be approximated. How often does the gauge spike? 350K miles @ 50 mph average = 7,000 hours = 420,000 minutes = 25,200,000 seconds. If the spikes are high and happen with great frequency (ie every couple seconds or more) then that bolt may be getting quite the workout.

 

[The patriot]

Something else to add to my repertoire.

 

[BigPapa]

Something else to add to my repertoire.

He said "repertoire".

 

[Wayne M.]

You might be on to something. How high do the pressure spikes go and what is the diameter of the open end of the filter housing? Once the force on the bolt from the spikes is calculated then the frequency needs to be approximated. How often does the gauge spike? 350K miles @ 50 mph average = 7,000 hours = 420,000 minutes = 25,200,000 seconds. If the spikes are high and happen with great frequency (ie every couple seconds or more) then that bolt may be getting quite the workout.

Not too sure about specifics, IIRC he said typical is #22- that sound right? I guess he's using a oil pressure gauge, and he said he's damaged and broke them. I don't see him often, and he said he lurks around on here. I hope he joins up.
It must be normal, because it was mentioned about the needle/seat and rubber hose to soften the impact on the gauge.

 

[GAmes]

For anyone feeling the need to swap out their bolts or carry a spare It is a 3/8-16 x 3.

 

[DavidC]

For anyone feeling the need to swap out their bolts or carry a spare It is a 3/8-16 x 3.

Thanks for the info

Dave

 

[The patriot]

Thanks. Will do.

 

[brods]

Not too sure about specifics, IIRC he said typical is #22- that sound right? I guess he's using a oil pressure gauge, and he said he's damaged and broke them. I don't see him often, and he said he lurks around on here. I hope he joins up.
It must be normal, because it was mentioned about the needle/seat and rubber hose to soften the impact on the gauge.

If it is really only 22 psi then that probably isn't enough to cause a problem. We've seen pressure spike peaks measured by an expensive fast response digital pressure gauge which were 50 times higher than what we saw using a standard pressure gauge, so its anyone's guess how high the spikes' peaks really are unless you have sensitive enough equipment to measure it.

Article about bolt fatigue here: http://www.asminternational.org/documents/10192/20564188/amp17208p18.pdf

The fastener assembly process is one of the most important, but often overlooked, contributors to fatigue performance. The cyclic-stress amplitude imposed on a given fastener (and therefore fatigue performance) is highly dependent on preload. Specifically, increased preload results in decreased cyclic stress-amplitude, particularly at loads below the clamping force imposed by the fastener. ... Thus, for a condition where cyclic loads are smaller than preload, little or no cyclic damage occurs to the fastener. The concept of higher preloads resulting in increased fatigue performance may be counterintuitive, which might otherwise suggest that increasing the tightening force of a bolt may increase propensity for failure. Insufficient preload has been attributed to the cause of the majority of fastener fatigue failures[4]. Therefore, understanding preload is critical in both the design and failure analysis of fasteners.

With the fuel filter bolt, if its not the cyclical fuel pulsations then the bowl must be side loading the bolt due to engine vibrations (or it is some combination of both). It would be interesting to watch the bowl with a strobe light to see if it was rocking or shifting side to side while the engine is running. If the low 10 ft/lb torque spec on the fuel filter bolt is due to the gasket design or limited by the strength of the bowl/mount assembly then the ideal bolt preload needed to fight fatigue may not be achievable. The torque charts estimate the torque required to achieve the proper bolt preload for plain, unlubricated, 3/8-16 bolts at 20 ft/lbs for a grade 2, 31 ft/lbs for a grade 5 and 44 ft/lbs for a grade 8. If the bolt can't be tightened as much as it needs to be then changing it out before it can fatigue seems to be the cheap and easy way to prevent inconvenient failures.



A bolt torque chart here: http://www.repairengineering.com/bolt-torque-chart.html