Bosch fuel pump problem

[RobH]

42 minutes ago, john.r.davies said:

If you have central heating at home with 10mm pipes, or microbores, you will notice they are fitted without insulation - adding that increases the surface area and so heat loss!

Er  - you might want to re-think that bit John.  "Ye cannae change the laws of physics".  The insulation would increase surface area but the added thermal resistance will reduce heat flow ands hence the losses. 

(Also heating  pipes are uninsulated because the heat is not 'lost' at all - it contributes to heating the house.  Insulation would be entirely wasted so cost ineffective.)

 

 

 

Edited October 15, 2023 by RobH

[Jules TR6]

40 minutes ago, RobH said:

1 hour ago, john.r.davies said:

If you have central heating at home with 10mm pipes, or microbores, you will notice they are fitted without insulation - adding that increases the surface area and so heat loss!

Er  - you might want to re-think that bit John.  "Ye cannae change the laws of physics".  The insulation would increase surface area but the added thermal resistance will reduce heat flow ands hence the losses. 

(Also heating  pipes are uninsulated because the heat is not 'lost' at all - it contributes to heating the house.  Insulation would be entirely wasted so cost ineffective.)

I know that silver (or white) surfaces reflect heat, black absorbs, that's a point against the black nylon pipes.

I know that metal conducts heat quicker than nylon, it also conducts cold quicker. So any heat absorbed directly over the engine (and it'll be less than the black pipes - won't it?) will tend to dissipate away through the rest of the metal braided hose - so increasing it's heat-sink value. That's another point against the black nylon pipes.

The metal braided hoses have a much larger external diameter, this might increase the surface area for heat-sink cooling - or it might increase the area to absorb heat! - I don't think there's a win either way here.

As long as the metal braided lines are not touching the engine, there is a possibility that all along the advice 'to go back' to black nylon lines has been misplaced. I'm keeping the s.steel brained hoses!

I do always order silver coils, for the same heat reflective reasons as above. Why would anyone order a black coil?! - for sure they don't look any better, and probably fail quicker.

Edited October 15, 2023 by Jules TR6

[Waldi]

4 hours ago, john.r.davies said:

I thought to investigate and educate myself a little more about this.   It's complex, as petrol is a mixture of compounds, all with different vapour pressures, but Lucas Pi operates at 105psi, when the "Reid Vapor Pressure" (measured at 38C) of the mixture is between 1 and 20psi.

The temperature at which a liquid will boil is defined as the temperature at which it's vapour pressure exceeds the ambient pressure, usually atmospheric.    That Boiling Point of the mixture that is petrol is 65 to 90C. 

 

This graph indicates that too, but a Lucas Pi system, after the pump, is regulated to 105psi, when atmospheric is about 15psi.   So the pump and post pump piping will need to be hotter than before the fuel boils, and produces bubbles of vapour in the system.  60C is more than 'hand hot' and 90 is scalding hot!   

Is "fuel vapourization" a false fear?   Next to engine, after a run and sitting in the sun, fuel lines in the engine bay might get that hot, but while running?

I'll be grateful for comments from others who know more than me!

John

Hi John,

the vaporisation pressure of a liquid is typically defined at atmospheric pressure (0 psi measured at a gage, hence 0 psi-g).
Petrol is a composition of several hydro-carbons and has a boiling range, starting at around 35 degr. C (at atm. pressure). 

The static pressure in the inlet of the pump is the outcome of:

1) the atm. pressure

2) minus the pressure loss over the tank-outlet/piping/filter

3) minus dynamic pressure (which can be ignored in this case) 

4) plus the static head (of the fuel height from tank fuel level to center of pump inlet).

This pressure is called NPSH-A: (Net Positive Suction Head Available, which is basically the available pressure at the pump inlet.
But that’s not all. The pump requires some extra pressure above the vaporisation pressure at a given temperature at the inlet, to avoid it will cavitate.
This is called the NPSH-R (Net Positive Suction Head Required).

The latter is pump (design) related.

Bottom line is the pressure at the pump-inlet shall be higher than the vaporisation pressure.

Waldi

[PodOne]

No expert with PI but I've read many a post! Personally I think the motivation to switch to SS lines is for the bling factor.

The fact that they are screwed into the MU will mean engine heat will conduct along there length and add to the radiant effect which will happen albeit perhaps at a slower rate than compared to the nylon which lets face it have lasted 50 years in some cars without issue so I stuck with the nylon. 

Same goes for any coil bolted to the block they will get very hot, mine was so a move to the inner wing which solved it. 

Andy       

[Mike C]

4 hours ago, Waldi said:

Hi John,

the vaporisation pressure of a liquid is typically defined at atmospheric pressure (0 psi measured at a gage, hence 0 psi-g).
Petrol is a composition of several hydro-carbons and has a boiling range, starting at around 35 degr. C (at atm. pressure). 

The static pressure in the inlet of the pump is the outcome of:

1) the atm. pressure

2) minus the pressure loss over the tank-outlet/piping/filter

3) minus dynamic pressure (which can be ignored in this case) 

4) plus the static head (of the fuel height from tank fuel level to center of pump inlet).

This pressure is called NPSH-A: (Net Positive Suction Head Available, which is basically the available pressure at the pump inlet.
But that’s not all. The pump requires some extra pressure above the vaporisation pressure at a given temperature at the inlet, to avoid it will cavitate.
This is called the NPSH-R (Net Positive Suction Head Required).

The latter is pump (design) related.

Bottom line is the pressure at the pump-inlet shall be higher than the vaporisation pressure.

Waldi

+1. The vapourization occurs with hot fuel in suction lines operating at low absolute pressure. HP lines after the pump will not suffer from fuel vapourization.

[Mk2 Chopper]

So would an under wheel arch pump fitment, actually be more susceptible to cavitation due to the length of the low pressure suction line, compared with a spare wheel mounted pump that has a shorter suction line? 

Edit: 

I'm thinking it's probably just mounted on the outside of the spare wheel housing, so actually not much further away. So would an addition of a small pump that can suck better and feed the bosch pump prevent cavitation?

Gareth

Edited October 16, 2023 by Mk2 Chopper
Correction

[Mike C]

You need to keep the head, basically due to the fuel level in the tank, on the pump suction as high as possible along the pipe run right into the pump suction. Large suction pipe diameters, low pump level, high fuel levels, low pressure loss suction  filters and straight pipe runs all contribute.

A spare wheel well mounted pump has a better chance of getting all of these characteristics right.

As I've noted in posts before I largely avoid the cavitation problem by having a Carter pump, which can lift fuel much better than a Bosch pump, mounted in the spare wheel well and it then pumps fuel into a Bosch pump mounted alongside the rear passenger side wheel arch.

[Mk2 Chopper]

Yes i remember now you've said that. What's the reason there isn't a better pump to combat this problem? Do you know the specification of the Carter pump you use? Does it require any regulation or is it quite low pressure?

Gareth

[Mike C]

The primer/ booster pump  was a very popular mod in the 1970's/80's .It's covered in Roger Williams older book " How to improve Triumph TR5,250 &TR6", chapter11.

My Carter pump is an ancient Carter Universal from what I can find, my greasy old notes indicate I was looking at  a Carter P 60504 as a replacement when I thought I had a problem with it.. But it doesn't matter , there are better pumps available today- Williams recommends a Facet Red Top with internal strainer.

Personally I'd select a carburetor engine pump with very good suction characteristics, a very nominal head  of say 5-10 psi, and a flow equal to or a bit more than than the Bosch unit you have fitted. For reference Williams says the original Lucas pump's rated flow rate was 72 L/hr.

My primer pump's  wired in parallel with the Bosch pump. There is no separate flow regulation needed on the primer pump, the Bosch pump and  the  PRV will set the flow for both.

Re reading Williams book after a number of years, it's apparent that his main concern back in those days with the original  PI fuel  system was, indeed, fuel vaporization.

 

Edited October 16, 2023 by Mike C

[Mk2 Chopper]

Thanks Mike, this is really interesting information, I might try get hold of that book too. Yes I see what you mean a carburettor pump would be ideal, going to have to do some research on pumps now! 

Thanks

Gareth

[Jules TR6]

Wasn't this what the Bosch fuel pump upgrade was supposed to resolve - that it ran much cooler than the old Lucas pump?

Fuel to and from the engine probably gets cooled somewhat on it's run under the vehicle.

Does anyone still suffer from cavitation in the fuel affecting the Bosch pump - or being caused by it?

Fuel lines above the engine becoming hot when a hot engine is turned off, is another matter of course. And even with all measures to prevent cavitation (i.e. having a 'pre-pump') I guess we'd still all want to park with the boot under shade on a summer's day.

Coils failing when they become moderately hot does seem to be a more common issue however.

[Mk2 Chopper]

27 minutes ago, Jules TR6 said:

Wasn't this what the Bosch fuel pump upgrade was supposed to resolve - that it ran much cooler than the old Lucas pump?

Fuel to and from the engine probably gets cooled somewhat on it's run under the vehicle.

Does anyone still suffer from cavitation in the fuel affecting the Bosch pump - or being caused by it?

Fuel lines above the engine becoming hot when a hot engine is turned off, is another matter of course. And even with all measures to prevent cavitation (i.e. having a 'pre-pump') I guess we'd still all want to park with the boot under shade on a summer's day.

Coils failing when they become moderately hot does seem to be a more common issue however.

In reality was it more a case of an alternative to the lucas set up when spares were not so easy to come by back in the day, or to solve the cavitation issue? I still hear of people that suffer with cavitation issues with the newer system, so it may have solved some (most) but not all cases. 

Very little fuel comes back from the metering unit, most of the heating is done at the pump and returned straight to the tank via the prv to go round and round getting hotter and hotter until you reach the threshold of the boiling point of petrol. Its the gravity fed suction side that's the problem here, hence the idea of another pump with better suction to feed the Bosch pump. 

Coils I'm not convinced about, why would a coil no longer work when it's hot, but work again when cool? Is it more likely a bad connection at the coil, and when swapping coils sorts the issue out? 

Gareth

Edited October 16, 2023 by Mk2 Chopper
Correction

[Mike C]

The Bosch pump has an inherent problem in that it can't tolerate much of a vacuum in it's suction line, it can't lift ("suck") fuel much without cavitatng. This gets worse if the fuel is hot.

A GP carburetted engine pump  is much better at  lifting  fuel  but can't generate  much pressure.

So the idea is to use a pump that's good at lifting fuel to feed a pump that's good at generating pressure. If the two pumps are reasonably well selected this gives a combination that does both .

Fuel pumps are such reliable components that overall system reliability is not a problem.

 

 

[Jules TR6]

18 minutes ago, Mike C said:

The Bosch pump has an inherent problem in that it can't tolerate much of a vacuum in it's suction line, it can't lift ("suck") fuel much without cavitatng. This gets worse if the fuel is hot.

A GP carburetted engine pump  is much better at  lifting  fuel  but can't generate  much pressure.

So the idea is to use a pump that's good at lifting fuel to feed a pump that's good at generating pressure. If the two pumps are reasonably well selected this gives a combination that does both .

Fuel pumps are such reliable components that overall system reliability is not a problem.

 

 

Very informative. This is going to get me checking the fuel tank outlet bore size and the straightness of the hose to the pump. I keep meaning to get back there to try and find out the cause of an occasional squeaky buzzy sound (often heard more on slowing right down - which makes me think it's the PRV or the return).

This will coincide with some research into an additional 'sucker pump' - potentially an upgrade for next year!

[john.r.davies]

The two-pump solution has been mine too, but more to deal with fuel slosh in the tank than vapourization.  I originally fitted Pi to my Silverback, the world's only race Vitesse estate!  The estate has a flat tank nearly the width of the car, with minimal baffling and a tiny, gravity fed 'pot' to combat slosh, and a swirl pot was the solution.

I use a Facet to keep that full, and now mount it as high as possible in the boot of Sb's successor.   

John

[Andy Moltu]

The temp charts were of boiling points at atmospheric pressure. 

Other factors to consider are the vanes in the pump itself causing the pressure inside the pump to be lower in places which will impact boiling point and are there local hot spots inside the pump mechanism in contact with the fuel?

As has been pointed out things like braided injector lines can result in vaporising in or near the injectors which compounds the issue.

Are these after market pumps really capable of supplying 100+ psi for prolonged spells. They may do 150 or more in testing but do they do long term tests at our pressures when the target market is 45psi or so. You might sprint a short distance uphill but not all the way up a mountain.

 

[JohnC]

18 hours ago, Andy Moltu said:

As has been pointed out things like braided injector lines can result in vaporising in or near the injectors which compounds the issue.

I think this may be conflating two issues. One is the fuel vaporisation upstream of the pump (well covered by others in this thread). The other is fuel vaporisation caused by heat soak along those braided lines with the hot engine stopped. When you try to restart there's a lot of injector line priming required. The clamps holding the injectors into the throttle bodies are the likely path of heat conduction to the braiding via the injectors, unless of course the PI lines are in direct contact withe the engine elsewhere. It's not about heat absorption by the braiding from the ambient under bonnet temp, it's heat soak from the cylinder head via a metal pathway.

Perhaps somebody could test this by grasping the braided line after a long run and a 10min rest? I'm happy to do the same with my original nylon lines...

JC

[Andy Moltu]

Both issues combine on a hot engine. 

A pump that is cavitating may well still be delivering fuel at a reduced pressure. 

If that pressure is enough to open an injector that's fine. If not the car stalls.

Compound that with an injector/line that is full of vapour rather than liquid you may well struggle with a restart. (Pump supply voltage dropping under cranking) The pressure delivered by the pump may well fall enough to fail to open and bleed out the injectors. 

If the pump can't deliver fuel at the required pressure, the PRV will be closed and the flow through the pump stops. The pump needs this flow to cool it. 

I may well be talking about 2 problems but they do interact and combine to an extent.

Braided hoses are perceived as more secure but that has to be put into context by considering the insecurity of the standard fuel lines. How many of those fail in use? Not many if at all. Damage tends to result from spanner slips when working on the injectors or metering unit.

Thus they probably should be largely considered as bling. There are many anecdotal reports of fuel vaporisation in them. The plural of anecdote is not necessarily fact but there are sound reasons to think why this might occur.

[JohnC]

On 10/18/2023 at 7:56 PM, Andy Moltu said:

The plural of anecdote is not necessarily fact

I only just read this post. I love that observation

 

[Casar66]

Here in Germany, too, a "holistic" approach to injection is always a major topic. Temperature of the fuel, pressure of the pump, swirlpot of the tank, diameter of the tank drain, double pump, cooling of the ignition coil, etc. So there are many possible reasons why a PI does not run properly, especially at higher temperatures. But at the end of the day, there is always the realization that there is no single cause or measure. And there are TRs that run perfectly well and others that are a pain. I have not yet been able to establish a connection as to whether TRs with a lot of additional technology and improvement measures run better than those that are kept quite simple. 
My CR, for example:
Just a Bosch pump instead of the original one, installed together with the pre-filter in the spare wheel well, has never caused any problems (except when a component fails technically, like the rubber membrane in the MU). Not in traffic jams in Genoa at 40 degrees, not in Corsica at the height of summer, simply never.
Since I don't believe in miracles, the cause of possible problems with the injection system must lie elsewhere than in upgrading a general working standard set-up. 

 

[JohnC]

Totally agree. There are many "improvements" that only improve the wealth of the vendor. Like an external oil feed to rockers (on a standard engine). Or an oil cooler for a daily driver in the UK. Or, just possibly, braided fuel lines... Sometimes it's best to make sure each & every component of the original system is working as designed.

[Andy Moltu]

Just bear in mind that the original system was cobbled together by Triumph pinching every penny they could to get the Lucas PI system down to the budget allocation! So often the original in factory condition was marginal and as a consequence small deterioration can be enough to mess things up.