Nissan Juke : Juke Forums banner

401 - 420 of 454 Posts

·
Registered
Joined
·
1,306 Posts
Looks so cool and well thought out!
 

·
Premium Member
Joined
·
628 Posts
Like the design and thought process.

Did you think about putting in some kind or pressure relief plug/valve in case of backfire? Could be as simple as a rubber plug designed to blow out in a backfire or major spike in pressure. You might be able to build something a bit lighter if you added one to alleviate the worst case scenario rather than building it to handle those pressures.
 

·
Registered
2011 Nissan Juke awd cvt
Joined
·
10 Posts
Mileage estimates are 100k plus. The forged pistons are gonna wear the bores more though. From my understanding engine warmup is mandatory, not optional on this setup. The upsides are worth it though. IMHO the Juke have a strongish stock piston but the rods are more questionable. I’m not taking any chances in case I get a boost spike or detonation. Plus, the forged pistons and rods cost maybe $500 more than the OE rods and pistons, not that much of a difference.

It’s very likely the 2J upgraded rods could be reused on another rebuild, maybe just requiring a piston swap. They are rated for 650 whp. The stock parts I don’t think that’s the case. The V1 engine blcok can easily handle 1 engine rebuild due to the thicker iron bore liner thickness.....possibly 2 rebuilds. The V2 engine block liner is much thinner. Once it’s out of spec it’s a throwaway block.
That 650 is a made up number. There is zero data backing that
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #404
That 650 is a made up number. There is zero data backing that
That is from 2J's website. Given the cross sectional area of the rods, 650 hp wouldn't be a problem.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #405
Congrats on the custom intake manifold. Looks like you have a pretty crazy CVT build going on yourself.
Start a thread and postup some driving impressions.

Cool to see there is more than one way to build a CVT.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #406
Update:

Quotes came back on the Duraform HST today with the latest design. Higher than what I expected but doable and about what a high end Intake manifold would cost from Blitz for my EVO X. I'm redesigning and negotiating to make it more reasonable and possibly more durable. Ultimately you gotta pay to play. The Duraform HST is weaker than the Windform SP. Still waiting on quotes from Italy but it will be insane expensive going thru a 3rd party. I kinda know who they are so maybe I go direct. I'll have to redesign the manifold slightly to survive using the Duraform HST on the runner section but not much. The plenum I have the option of running an off-shelf aluminum plenum casting with a custom CNC machined base plate and throttle body flange. It'll look almost identical and functionally the same once powder coated. Still cost as much, but now the aluminum will add some durability as well.

This'll be a process, but it's going to be done right.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #407 (Edited)
And.......here we go:




188761
 

Attachments

  • Like
Reactions: squirtbrnr

·
Super Moderator
Joined
·
1,778 Posts
Science!
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #410 (Edited)
Lol.......this is science.

Another look at the port velocity.

188764
 

Attachments

·
Super Moderator
Joined
·
1,778 Posts
so is there a problem with the difference in relative velocity between the ports? Maybe not at higher boost levels, but could this make maybe cylinder 1 (furthest from throttle body) always run lean and cylinder 4 (closest to throttle body) always run rich? Just thinking about air being forced into a cylinder and anther being starved due to velocity of the air.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #412 (Edited)
so is there a problem with the difference in relative velocity between the ports? Maybe not at higher boost levels, but could this make maybe cylinder 1 (furthest from throttle body) always run lean and cylinder 4 (closest to throttle body) always run rich? Just thinking about air being forced into a cylinder and anther being starved due to velocity of the air.
Yes, it's a problem. Exactly, furthest cylinder gets rammed with more airflow.....basically it's intuitive. The airflow can't make the inside turn without recirculation and lower pressure. That little bit pressure difference from inside turn to outside turn really messes with the flow balance across to the ports. HVAC systems use turning vanes........but there is a much more elegant solution to the problem without adding any measurable restriction or complexity.

I'm not showing it for now but I'll show the data below, though I'm still perfecting the geometry. I'll say right now this could not be done without a CFD computer simulation or 100's of hours on the dyno. Similar to extremely advanced racing intake systems but my own design for fitment inside of a compact street plenum. Theoretically, it should work and the CFD confirms it's having a dramatic affect. It's gonna be baller though.

Tuning:
Now from a tuning perspective, which is where I'm going with this, you can tune to the edge of blowing up the motor and simply outgun your competitor on the same setup. This concept get's totally lost on guys who aren't experienced in ECM tuning, not saying I'm an expert by a mile. But a professional tuner understands how important that aspect is. For now, let's just say cylinder flow balance is the first step to "intake tuning". Get that right, then can move on to tuning the intake geometry for the respective rpms, torque, power goals. Honestly, I have to run Dynomation-7 to do it but it will very accurately park the numbers where a tuner wants them to be and also provide a dyno curve as well. From my understanding, it get's you very close to the final geometry with some final prototype tweaking.

The flow imbalance looks worse than it is actually. I'd have to run a true pulse/time based CFD simulation to show the difference in mass flows, that's what's important. In reality, it might be +/-10% difference in mass flow from cylinder to cylinder, best to worst for example. The velocity profile is just to show the relative differences and a quick gut check. Tuning wise, even +/- 10% without a power robbing tune to compensate.

Manufacturing:
I got insanely lucky yesterday and came across an outfit in Michigan that do nothing but Rapid Prototyping for the racing industry. They do stuff for IMSA, Indy Car, Ford, Porsche. The guy talked my ear off about fabrication techniques. They are masters at doing SLS intake manifolds, welded aluminum manifolds, everything. They go on racing cars and OEM vehicles for development for the big boys. They are in Michigan, so if you understand what that means........these guys do car stuff and now it. I quizzed him on some engineering concerns I know about, he knew everything I was about to ask about. I was looking to partner with someone, these guys may be it.

Anyway, we talked material selection, hybrid design concepts, o-ring vacuum/pressure seal requirements, surface finishing, heat conduction, manifold heat dissipation, backfire mitigation, everything. I'll make a spec list requirement here once I have time. These guys make them and know how to make them to survive, this was my biggest concern. Couple tricks to make them last on a street car cause we all like to think we are racers.......but we are not:) I prefer an aluminum casting, but this is what I can afford. I'm looking at a split runner aluminum design for 3-4 axis CNC machining as well, and so I'm reviewing that. I'm not opposed to welding billet aluminum pieces together like a welded split core injection molded design. I like a manifold to look like one, so I'm not designing a block of aluminum with runners cut into. A 3-4 axis CNC with some climb cutting or mold die cutting software would make short work of a 6061 aluminum billet. The 5-axis CNC quotes were astronomical. Engineers make a design fit the manufacturing process, so here there are a few options for future.

Revised Intake manifold design:
OK, so below I'm showing my "Rev 2" intake manifold design with my PORT BIASING technology added. Not perfect as it's taking a large number of permutations on the CFD to dial in, but my first pass radically swung the port velocity/mass flow biasis back away from #4 cylinder. I proved the concept. I'm not showing the design just yet as I'm tweaking it still and I have maybe some plans for patent application. I don't think it'll affect any Helmholtz or resonant tuning much but the flow balance will be radically improved. The goal is simple & compact but totally effective at balancing the cylinders to +/- 2% mass flow for example.


Velocities are ball park closer, still working on it. Pressure drop is unaffected. Got some more tricks to run on the CFD and I should get it closer.
Enjoy

188766
 

·
Registered
Joined
·
1,270 Posts
Gene - You are a mad scientist haha.
 
  • Like
Reactions: squirtbrnr

·
Registered
Joined
·
1,485 Posts
Discussion Starter #414 (Edited)
Lol, I try Storm Trooper. Mad Engineer though:)

And, I was able to balance the Intake manifold to within +/- 5% @ 40 lb/min (0.303 kg/s) or 400 h.p. The CFD is able to show the actual mass flow to each cylinder.
The dimensions were fairly critical to achieve this, probably need a supercomputer to get better results running an iterative solver. I tried many combinations that put the cylinder balance all over the map like +/- 50% which is insane before I hit on the right combo of design elements that worked. It's almost as good as a center mounted throttlebody manifold with a huge plenum, but it'll fit in the stock location without changing the intercooler pipe routing. I'm playing around with the final plenum shape to clear the engine block and make it a little cleaner looking.

Basically: The trick is the inside of the intake manifold is shaped like a turbo compressor housing. This diffuser plate or flow biasing orifice plate helps to feed out the airflow evenly, biasing more towards the 2 front cylinders, and throttling the rear 2 cylinders ever so slightly. This could not be done without the CFD software or an airflow bench. It took another trick at the throttlebody bend to decelerate the airflow and give Cyl #1 a chance to breathe, otherwise it all doesn't work. I simply used a diffuser or expanding pipe to slow the air down going around the bend so the momentum was reduced efficiently. Total pressure drop is only 0.25 psi from cylinder port to throttle-body, almost nothing.

Anyway, there you go.....a bit of science.



188771
 
  • Like
Reactions: redBack_juke

·
Premium Member
Joined
·
628 Posts
I like the iterative approach. As you are clearly aware, product development is a sometimes painful process of finding and addressing all the areas of improvement, not doing something once and declaring it perfect. Anyone trying to innovate taking a scientific approach gets to be wrong a lot about their assumptions throughout simulation, testing and validation- and the more advanced the application, the more often you get to be wrong. While some of us have learned this lesson the hard way, some people who have egos invested in always being right should remember that rigorously seeking out areas for improvement and figuring out ways to address them is how progress is made. The community overall benefits overall by you taking them through the optimization process in detail... I am fascinated and would love to see more of the iterations that didn't work, as you tend to learn more from them than the final design.

A few more thoughts:

As a further tweak, did you try individually adjusting the angle of the openings of the velocity stacks to the flow? The beauty of additive is that you can play around with designs that go beyond what is possible with traditional manufacturing.

Also, did you think about that pressure relief plug? Reducing the amount of material, as the worst case scenario is relieved, may reduce your costs, in addition to weight.

And while you are at it, did you think about a secondary injector into the plenum? It has been done before to add fuel flow or methanol on boosted engines and might be a consideration, depending on your end goals.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #416
Thanks Ine937s.

One thing I tried was venturi restrictors on each individual bellmouth to each runner. Oddly, it didn't work that great and was crazy sensitive. Mainly because you need to gain control of the airflow momentum well before the runner bellmouth opening.

The physics dictate the design, this is how I solved some critical balance problems. The air is like a jet stream with different velocity strata/layers and a lot of momentum. What you really want is all airflow vectors at the same speed feeding the runners, then they will get even mass flow. The air can be accelerated or decelerated using nozzles or diffusers while minimizing pressure losses. If Cyl#1 is feeding off an inside radius turn, naturally that air is slower just based on physics. Nothing you can do if the runner already is feeding off a low velocity stream, except restrict the other runners too, which is poor design. The trick is so to slow ALL air going around turns "first", trading speed for some pressure boost which is mostly recoverable momentum, turbos do this. Reduce ALL airflow speed feeding the plenum, and the velocity vectors feeding each runner are now much more balanced and the job is mostly done for you. Like turning a car at slow speeds, much easier.

The flow balancing orifice plate fine tunes things to rebalance the cylinders to a high precision with just the slightest restriction to Cyl #2, then a little more to Cyl #3, then even more to Cyl #4. It's like corner balancing a racing car, just a smidge on spring height adjustment shifts the vehicle weight around. That is what makes the whole thing work. The plenum inlet diffuser is totally critical to this runner balance, the orifice plate dials it into higher precision runner balance. It worked on the 2nd try when I followed that approach of optimization, a computer iterative solver could nail it dead-on.

I didn't try the velocity stack angling, but I will try it. I like the idea of angling them inside the plenum and not on the runners as it's too much redesign work. AMS try some funky side tilting bellmouths, like tulip flowers, not sure how well that worked. Cosworth do some bell inlet tilting as well. The forward tilting makes sense as it's more of a "turn" into the main flow airflow streams, yes I think it's even better than straight as I have it. Those little nice details simplify and make it more stable for varying conditions. Yes, I like your concept.

The relief valve is called a "burst plate", it's used on NHRA funny cars. I designed a PRV for a water softener, so I'm familiar with the concept. It's kind of like a blow off valve, just set at a higher pressure rating. I forget the rating but those things run 60 psi boost anyway. It's resettable with some type of silicon diaphragm., I might just buy one off the shelf and design the mounting pad on the intake manifold.

Extra fuel injector bosses I didn't even think about, another thing I could easily add. I would have to precision drill the material to get the o-ring seal perfect, but otherwise yes I could add them and plug them off. Good suggestions.

I'll add those items and see how it looks.
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #417 (Edited)
OK, what I'm about to show is a dual manifold design from Audi, they used this on their Rally cars circa 1983--1988 and AUDI R8 lemans cars. They are supposedly designed to flow balance the runnners. HKS made one for the EVO 1-9 as they stole the design from Audi. But they didn't make great high rpm power......and I can see why that is now. Stephen papadakis tried this style manifold on their current drift cars. But on their +1000 h.p. Supra they went back to a more conventional intake manifold design which was 3D-SLS printed in aluminum. I understand now why.

Anyway, using this Audi dual plenum design I cannot for the life of me balance this manifold design. The amount of swirl it creates is intense, the pressure drop is about 2.0-2.5 psi which is very high. That would be equivalent to losing about 6% power output or at 400 h.p. you would lose about 24 h.p. from the pressure loss. This is extremely bad.

This design looks awesome in theory, on the CFD it's a hot mess express for flow distribution. You want to fire the flow without altering the momentum, keep the flow in the same plane otherwise you start rotating/swirling the flow which is very extremely bad. There is a reason this manifold design fell out of fashion.

I'm working on making the flow balance orifice plate removable as it's sized for a specific peak horsepower.


188783
188781


188785


188786
 

Attachments

·
Premium Member
Joined
·
628 Posts
So, while I was reading this, my wife was vacuuming and making it hard to concentrate. But then looking at the dust separator on the vacuum, and it gave me an idea. Below is a VERY rough 2D sketch, not to proportion/size, etc.

In the dual manifold design, air goes from the throttle body, to be spread out in the first manifold, then sped up through a restriction, then slowed down in the second plenum, then forced to change direction and speed up again through the runners. This may distribute the charge, but the associated losses present a challenge...

But what if you used that high velocity air to create a vortex inside a cylindrical plenum, with the runners pulling from that vortex? Take that tendency to rotate/swirl that would otherwise be bad, make it intentional and amplified and put it it use. Like in that dust separator on my wife's vacuum, pressure would increase along the outside edge. Rather than having velocity stacks enter the housing to accelerate the air into the runners when the valve opens, you could have adjacent intakes along the outer edge of the vortex pull high speed air into the runners. Potentially using a modification on a NACA duct or other low-drag intake design to maintain velocity in the plenum. Using additive, you could also have greater integration of the design that naturally would wrap around a plenum in the center.

Obviously, it would take some work to get the design right. Nozzle position, shape, intake position and shape, size of the plenum, etc., would all need to be figured out and optimized. You would need to maintain the vortex even though each valve is only open 1/4th of the time, so sizing would be important.And you would likely want a different shape of runners, etc., to make the most of the flow... But what do you think?


188791
 

·
Registered
Joined
·
1,485 Posts
Discussion Starter #419 (Edited)
Ine937s,

Awesome, we think alike. They plenum taper is achieving the same effect of making a big circular barrel, in this case a narrow manifold. Cool idea and very much along the same lines but I like the innovation. I still can't believe Audi are using that side feeding design, it doesn't work good on the simulation. It looks awesome, but looks are looks.

I have my design/invention using venturi bellmouths to accurately balance massflows thru the runners. The total restriction in the manifold is about 0.1 psi with these metering flow devices. Looks a bit weird but there are ITB setups that use 25mm dia throttlebodies for some road racing turbo setups. This gives you an idea how much opening you really need at higher boost levels on a turbo car and still make big power......not much more than 1" at each runner. I'm running about 1.625" but can drop down to 1" using a bellmouth/venture/diffuser and have almost zero restriction, or enough to just balance the runners. Couple of other tricks but a 90* turning vane aft of the throttlebody is becoming critical to feed Cyl #1 in a tight turn. I'm also using a 2.36" to 4" diffuser to get into the plenum efficiently. Again, 400 h.p. capable, cylinder balance within +/- 2.5 %, and nearly 0.1-0.15 pressure drop. I should be able to fabricate a compact intake manifold in the same spot as the stock manifold.

My Design elements are as follows:

1) Any design elements Skunkworks are using as mentioned, which I've found by accident but also by inspection, too many to list here for now but have been simulated to work effectively.
2) Turning vanes in the throttlebody plenum adapter. Required for our Juke's 90* throttlebody entry requirement, custom machined and bolt-on to the plenum.
2) Flow balancing venturi bellmouths in the plenum runner inlets with varying sizes and lengths for specific tuning requirements.

These (3) things greatly helped on simulation to reduce the pressure drop to about 0.1-0.15 psi total from throttlebody to cylinder head. The bellmouth venturi help alter the extremely tiny pressure drop differences across the runners to fine precision so the airflow is perfectly splitting the flow 4-ways to the cylinder head. This cannot be done in the plenum despite what anybody claims. The trick is to put all (4) runners close to each other so they sense similar plenum pressures, the differences then end up being much less, I'm still working on that but I've tried a few runs and it works nicely. Having the plenum designed correctly and the runners up close to each other and in the right "group" position within the plenum will make the bellmouth venturi flow controls easier to manage things. I'm feeling the venture may make Helmholtz tuning tricky but I'm giving it a shot anyway.

The goal is to use the largest venturi orifice size needed as to not pose a problem or interfere with the Helmholtz pulse waves. I've already simulated the runs using exact venturi sizes and got about +/-2.5% runner balance without huge effort. The restriction to the engine was essentially zero. It's horsepower dependant, so they must be re-tuned for a range below 300-400 h.p. Lower requires different sizing but it's only a tuning adjustment and then the balance is back again. Individual A/F readings in the exhaust manifold will be needed to see what's going on.

Manifold construction:

I'm redesigning the manifold using offshelf aluminum castings, custom CNC machined parts, and welded tubes. Cost is the same, reliability bulletproof. I have it costed out and parts alone maybe $900, labor yet to be quoted. The main issue is backfire and heat. I did a long engineering analysis on the temperature curves for this Duraform HST rapidprototype material vs. PA6-GF30 used in production. The results were not good enough for surviveability and cycle loads or run hours, I had all the data from the suppliers. Once I started de-rating everything it just didn't make sense for the money involved. Prototype, great......but for long term never would it survive. For the money, the method above is better and bulletproof.

Redesigning over the next few weeks and will post up pics of the final machined/welded/cast design on the computer. Honestly, this is the best way for a (1) off manifold design.
 

·
Premium Member
Joined
·
628 Posts
From a construction perspective, did you happen to look at the Virginia Polytechnic Formula SAE team? They used FDM 3D printed parts for the shape and smooth inside surfaces. Then they wrapped it in carbon fiber and vacuum bagged for strength. Granted this is a small engine, but it seemed to work well. This was 2009 and additive has come a long way, so I am sure something better has come along. But a multi-material approach, using easier to come by additive manufacturing for the smooth internal surfaces combined with a strong composite outer layer might be something to think about, if you haven't already.

 
401 - 420 of 454 Posts
Top