VTT N54 HPFP Information/Disassembly/Upgrade Post

Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
I finally had a few minutes to sit down and take a couple pictures of some old HPFP's and share with the community. The goal of this is primarily to take some of the magic out of how this particular HPFP works, show how the unit itself *could* be upgraded if you had the time, money, and desire, and also show why we (VTT) decided to just spin the damn thing faster.

First, lets look at the beautiful, albeit dirty HPFP connected to a vacuum pump:

full.jpg


Oh, whoops, was that a VTT hat in there by accident? Almost looks like an advertisement!

Removing the HPFP from the vacuum pump is simple. Resultantly you're left with this marvel of modern engineering:

full.jpg


Here is the control valve:

full.jpg


When you first pop open the HPFP, you'll be met with a loss of fluid. How much? Top secret, which is code for "proprietary industry info", which in this case also means I forgot since it's been years since I pulled it apart. I'd recommend weighing the assy before and after disassembly and calculating fluid weight/volume the old fashioned way. I digress. This is, sort of, what you'll see:

full.jpg


I say "sort of" because it's in a state of disassembly here; you'll see three pistons with "shoes" that ride on the swash plate (angled thing on the left side). You see the triple chambers of the pump here; each get their own piston/shoe/spring/etc. which we'll go into more in a bit. First, though, how to disassemble them all easy like I show? Make a tool if you don't want to do it like a savage:
full.jpg


It's easy as pie to take them out now:

full.jpg


I've heard people prattle on about how the HPFP has to be assembled in a clean room and no one can ever even peek inside or the wizards will fly out. Little did you know, cat fur is an important ingredient. Supporting fuzz provided by Walter the Metric Cat:

full.jpg


Let's get into how it actually works.

full.jpg


Disregard the heat marks on that swash plate; that is a long story that involved some stubborn disassembly and a torch that was near by. Decisions were made. Yours shouldn't look like that.

So looking at the above picture, you see the swash plate turns when driven by the engine. This causes the pistons to move in and out, flexing the bellows with each pump stroke. The bellows fit nice and tight within the actual pump cavity and it simply takes up more or less space. You'll also see that the cavities have, in general, one fuel inlet/outlet. One hole does it all, so to speak.

full.jpg


Do to my less-than-stellar photography skills it's hard to see. Lets see if we can see better with terrible brightness and MSPaint:

full.jpg


Here are the "magic" parts assembled:

full.jpg


Quick rundown of how it works:
-piston/shoe is at thinnest part of swash plate
-fuel rushes in between cavity and bellows
-some minor compression of bellows occurs -not significant
-swash plate rotates
-piston/shoe ride swash plate and get driven in towards pump
-bellows is forced to expand
-during expansion of bellows, the volume in the chamber decreases; this causes a drastic pressure increase (fuel isn't compressible)
-check valve that let fuel into camber closes
-fuel is forced to pump outlet under significant (we hope) pressure

Simple as that.

Remember though, there are 3 of them. I haven't pressure tested the cavity to see where in the internal plumbing the check valves are relative to each chamber; doesn't matter for our purposes.

Let's have a look at a few basic measurements. Note: Despite using calipers/depth mic, these are not precision measurements, they were done with one hand while taking a picture with the other; the goal is to give you a general idea of size, don't go engineering something based off these numbers and get pissed at me when it's off 20 thousandths.

Depth mic to inside lip of bottom of hole:

full.jpg


Thickest part of swash plate:

full.jpg


Thinnest part of swash plate:

full.jpg


Thickness of piston guides:
full.jpg


Whoah whoah whoah!!!! That isn't stock. Add 0.040" to that measurement for stock thickness. It's almost like mine are custom machined for some strange purpose. Hmmm. Anyone want to guess what that was?

On a pump like this, that flows due to compressing/expanding the bellows, you could increase flow capacity a few ways:

1: Increase diameter of bellows. As you can see... not much (any) room for this.
2: Increase compression of bellows. Doable.
3: Increase pump stroke via upgraded swash plate. Doable.

I almost immediately wrote off the idea of getting a custom bellows made; it's expensive. If the quantity were sufficient it'd be worth looking at it but that still doesn't increase pump capacity by itself.

What I did was machine 40 thousandths off the top of the piston guide, and install 40 thousandths of washers between the piston guide and the bellows; this'll let it compress a little more before being smashed back into the HPFP on the compression stroke. It's not a lot but should help. It'll also make the HPPF respond a little better to higher LPFP pressures. Hmmm.

You could machine the pump down a small amount, change swash plate to one that throws more, but then you need a custom bellows, or just hope the bellows are ok with the extra compression/expansion... how much would that help? 20%?

Or you could spin it faster like the diesel guys do and the modern cam driven HPFP's are -put a 4 lobe cam on a HPFP designed to be driven by a 2 lobe cam and you effectively oversped it by 100%.

So you all have enough info to conjecture and play design engineer. Ask questions -I'll answer. I have several methods that will absolutely 100% work to make this pump output more volume and pressure, but... none of them are really practical/cost effective compared to the shotgun from us or even port injection. It's doable as a one-off pet project but not really as a production piece without significant investment. Unless of course one of you come up with a better method/idea, in which case more power to you and I'll be the first to congratulate you.

Our take? Enter the VTT shotgun.

Chris
 

doublespaces

Administrator
Oct 18, 2016
9,206
4,195
0
AZ
Ride
2009 E93 335i
I'll buy a stoked hpfp, paypal ready.

Is it more a design or labor cost limitation? Or would the cost of the actual upgraded parts be prohibitive? What kind of volume would be required to make one of those options feasible?
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
All of the above. Design is cheap when it's some joe fooling around on their computer after their normal job, but to pay a real design engineer that can actually give you results, not some bullshit ideas sketched in CAD (I don't know any.... ahem) it costs boatloads of coin.

Pretty much everything is fairly simple except the bellows. Getting something like the bellows made will not be cheap. It's custom, very thin metal, needs to be exact and withstand what you want out of it. You guys want me to destroy one and show you how that's made? I can post pics (when I get around to it). Testing it also would be, um, interesting since you all can guess what happens when it fails.

The actual machine work won't be too bad, but then you really need to be diligent about getting metal shavings out of it, since pretty much any contamination will kill you in short order.

If you told me you'd buy 10,000 units I'd look into it, I'm not sure I'd bother for less. At that point it would still a development program, whereas spin the damn thing faster absolutely works. If you disassociate emotion from it the choice is clear.

Back to this HPFP, post your questions, I'll rip this B apart and show you its guts. Just be patient if I'm a little slow to respond... I have a lot of other projects I'm working.

Chris
 
  • Like
Reactions: [email protected]

Bmwfixerguy1

Lieutenant
Jun 5, 2017
873
288
0
Ride
07 335i MT
Very good write up Chris. I did the exact same as you. Took it apart and then saw what needed to be done. Realized I couldn't do it and then realized even if I could how much would it cost for what it's worth.

One thing I thought of I retrofitting to an n55 style pump and and trying similar tricks that vag guys use that you mentioned above
 
  • Like
Reactions: [email protected]

Reaper0995

Specialist
Jan 10, 2017
69
20
0
Few things. First, can the smash plate have more than one lobe? Rather than being one big angle, make it a half pipe shape?

Second thing, I may have someone who can make the bellows and weld them as you have shown. I'm a manufacturing engineer and have a lot of suppliers that can do some pretty tricky stuff, Andy this looks like something I've seen before believe it or not. I do remember that's the bellows and welding were fairly pricey though, so even if they can it may be fairly pricey.

Third idea, could a new housing be made that would be large enough to house four belows or more?

Last idea, could the chain and shaft take more load? Would it be possible to put a 'spacer' between the pump and the engine that has a cam lobe driven pump? Making a housing and a single lobe cam wouldn't be terribly difficult, Andy the housing could hold one or two pumps like many OEMs do.
 
  • Like
Reactions: doublespaces
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
Few things. First, can the smash plate have more than one lobe? Rather than being one big angle, make it a half pipe shape?

Second thing, I may have someone who can make the bellows and weld them as you have shown. I'm a manufacturing engineer and have a lot of suppliers that can do some pretty tricky stuff, Andy this looks like something I've seen before believe it or not. I do remember that's the bellows and welding were fairly pricey though, so even if they can it may be fairly pricey.

Third idea, could a new housing be made that would be large enough to house four belows or more?

Last idea, could the chain and shaft take more load? Would it be possible to put a 'spacer' between the pump and the engine that has a cam lobe driven pump? Making a housing and a single lobe cam wouldn't be terribly difficult, Andy the housing could hold one or two pumps like many OEMs do.

1: First, it's a swash plate, not a smash plate, but could it be done? Not really, it's a nice smooth transition, it'd be considerably more abrupt if you did that, plus the shoes on the pistons would have to have more flexibility -meaning more wear/shorter life/etc.

2: Oh sure, I know it can be done. Anything can be done, the damn thing was made in the first place. It's cost/time/MOQ that will bite you.

3: Absolutely impractical. Again, we put a man on the moon, we *can* do anything, but this would essentially require the entire pump be re-designed just using the OEM as a guide, and you'd still have to fit it in the envelope we have (things in the way).

4: That's a doable idea, but again you're dealing with space constraints, and you still have to deal with control. The plumbing is always doable but it's not completely trivial. If you really want to add one of those single cam-driven style pumps I'd just ditch the OEM HPFP in its entirely and run a bank of the less efficient pumps all in parallel for adequate flow, perhaps using a control valve on the inlet to the pump banks and run them all wide open. Again, doable but for what? You still have to "overspeed" the little cam pumps to get anything meaningful out of them, unless you run 4+ of them, at which point plumbing, control, and cost are going to be very high. P/N 13518604231 is $1082 from ECS tuning. MINIMUM two required, and you're still going to be under the capability of a standard HPFP that has been spun faster.

On our cars, if you ditched the vacuum pump and/or located it somewhere else you'd have more room to play, but then you have other problems to solve.

Chris
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
Very good write up Chris. I did the exact same as you. Took it apart and then saw what needed to be done. Realized I couldn't do it and then realized even if I could how much would it cost for what it's worth.

One thing I thought of I retrofitting to an n55 style pump and and trying similar tricks that vag guys use that you mentioned above

Thanks! Again, it is doable. It's just a hard pill to swallow when we have other options readily available for DI, and if you want to wash your hands of the whole problem, PI is effective and inexpensive. Granted, we like DI better, but thems the choices we have -that make sense- right now.
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
Also -just to be clear, nothing in this post is new. I started working on this project with @[email protected]! back in... what, 2013 or 2014? I don't remember, but it was before Fuel-it was an official thing. Anyway I took lead on this, we pulled things apart came up with designs that we both realized were impractical for this platform, and decided to move forward with other projects (better LPFP's/TBI/etc). At that point PI spacers were just coming out, Fuel-it! was turning into a big deal, and it was around the same time that Tony came up with the shotgun system. Steve went one direction, I went another and ended up working with Tony a year or two later. My point is that this is nothing new/special, but every time someone asks me about just upgrading the HPFP itself I feel this impulse to share some of this knowledge/history. I know Justin (Twisted) has something going on, I asked him to shoot me an email if he wanted to discuss our findings back in the day/give him my old documents/scans/tools, as I'm all for it if someone else wants to run this to ground, but I never heard from him so I finally decided to just post some of it here publicly.

Personally? I want to say I'm uninterested in upgrading this old hunk of metal but that's not completely true, there is 10% of me or so that has a passing interest in it. Old habits and pump knowledge from my submarine engineer days I guess. I am, however, willing to let one of you guys be the hero if you're so inclined.

I have been contacted by a vendor who has some interesting ideas for an alternate pump(s) config for the N54... but that's low priority with all the projects I'm running -again, we are able to hold 3000+ psi to redline with 750+ whp on 100% E85 on the double barrel shotgun, so out of necessity other projects take priority.

Chris
 

Bmwfixerguy1

Lieutenant
Jun 5, 2017
873
288
0
Ride
07 335i MT
Thanks! Again, it is doable. It's just a hard pill to swallow when we have other options readily available for DI, and if you want to wash your hands of the whole problem, PI is effective and inexpensive. Granted, we like DI better, but thems the choices we have -that make sense- right now.

I'm almost done with my PI if that says anything to the lucrative ness vs my wallet resources and ability factor lol
 
  • Like
Reactions: [email protected]

langsbr

Captain
Apr 5, 2017
1,219
707
0
Ride
07 335i 6MT e90
As the original HPFPs were prone to failure, what is the root cause? Do the bellows collapse or something, and not allow proper pressure?
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
As the original HPFPs were prone to failure, what is the root cause? Do the bellows collapse or something, and not allow proper pressure?

Root cause? Did you just come out of a corporate meeting? :p

If you want to see the RCCA, contact BMW, I'm sure they did one. We sure didn't, but we have more pieces of the puzzle than they did; we knew they failed, we knew the symptoms of failure, AND, we knew what they did to correct the issue. It's a little easier to fill in the blanks like that.

Newer HPFP versions include a magnet on the side of the cover that the swashplate/pistons/shoes are present in. I'll let the community figure out why.

Chris
 
  • Informative
Reactions: doublespaces

langsbr

Captain
Apr 5, 2017
1,219
707
0
Ride
07 335i 6MT e90
I'm a noob and have yet to have my hpfp fail. I don't even know the full symptoms of a bad one - car won't start/limp mode only? I'm assuming if it goes to limp mode so you can get to a shop but under no boost, it would mean it just can't produce any high pressure. I was curious as to what inside of the pump fails. Wouldn't knowing that help in ensuring that any modification for increased pressure/flow wouldn't be exacerbated?
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
Depends, the old school ones would just flat out die on you, no start, nada. You were out of fun. From what I've seen on the newer ones, they don't fail all that often, when they do it's after quite a few miles (usually), and even then they still work ok, just start losing some capability. The original ones (early 2007) failed with nearly comical regularity.

Chris
 

langsbr

Captain
Apr 5, 2017
1,219
707
0
Ride
07 335i 6MT e90
Interesting- I had read (never experienced yet) that the benefit of having the 2 fuel pumps was so you could limp it if the high pressure fails. If its a full no start - that almost sounds like a blockage or seizure of the pump? You also mentioned a fluid that spils out - is that a special lubricant or just regular oil?
 
Oct 24, 2016
1,152
1,189
0
43
Scottsdale, AZ
I didn't do an oil analysis, but for someone doing this as a serious long term project that would be a worth while endeavor... or just put some high quality oil in there and see what happens. Remember, our tranny's/diffs are "lifetime" oil too, lol.
 

langsbr

Captain
Apr 5, 2017
1,219
707
0
Ride
07 335i 6MT e90
Root cause? Did you just come out of a corporate meeting? :p

If we circle back, we can pivot to ensure that we see a paradigm shift in the pump. I don't currently have the bandwidth to take the lead on this, but if we follow best practices, we can disrupt the ecosystem and build consensus that rather than working in silos, and don't get push-back, the take-away will be next generation sweet spot in fueling. I don't want to get our lines crossed, and appear risk averse. We just need to make sure this is a strategic fit for the community so that we leverage synergies across all vendors. My plate is full right now, and this might be your core competency, so I empower you to move the needle, as this project has lots of moving parts (figuratively and literally). Just be sure that what you do is scalable, and that you think outside the box after getting all your ducks in a row and drill down into the nitty gritty of the issue. At the end of the day, it is what it is, we just don't want to miss our window of opportunity and have to punt.

I'll be out of pocket for a bit, so we should take this offline.


Sorry - WAY too long in a corporate world! :p