User not logged in - login - register
Home Calendar Books School Tool Photo Gallery Message Boards Users Statistics Advertise Site Info
go to bottom | |
 Message Boards » » arghx's technical ramblings thread Page 1 2 [3] 4 5 6 7 8 9, Prev Next  
arghx
Deucefest '04
7584 Posts
user info
edit post

^ I'm sure all that information on hood scoop efficiency is sitting somewhere in some wind tunnel test reports, and in past computational fluid dynamics simulations.

[Edited on October 12, 2011 at 2:31 PM. Reason : page 3 ]

10/12/2011 2:31:15 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

yeah i just don't know where to find it.

10/12/2011 3:29:25 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Here is a diagram showing a typical valve lift curve on a standard piston engine with fixed valve timing and lift:



Notice the valve clearance line near the bottom. Also note the drastic changes in valve acceleration as the valve opens and closes. On a basic continuously variable valve timing system the intake valve will open earlier, increasing the overlap between the two strokes. Now, compare this to a rotary engine with two intake ports and exhaust ports located on rotor housings, like a 3rd generation Rx-7:



A rotary engine has no valves and airflow is controlled only by the opening and closing of the ports. It's a little-known fact that rotary engines have had variable "valve timing" pretty much since day 1. The secondary ports ("P+S" above) can are staged based on throttle position on older rotary engines and based on rpm in the Rx-8 Renesis engine. Ports on a rotary open and close faster than the valves on a piston engine which creates especially strong pulsation effects in the intake and exhaust manifolds.



Another interesting thing about a rotary engine is that it has a compression stroke every 360 degrees of crankshaft rotation, instead of 720 degrees on a piston engine. Thus the ports are open much longer than a comparable piston engine. Careful use of port timing dynamic effects allows great volumetric efficiency.

[Edited on October 13, 2011 at 6:45 PM. Reason : 360 degrees]

10/13/2011 6:35:49 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

For those of you doing custom plumbing on a project...



One thing to keep in mind is that the AN size is actually in 16ths of an inch. So a 4AN line has a nominal size of 4/16", or 1/4". A 10AN hose is 10/16" or 5/8".

In my firsthand experience, the larger the AN hose the easier it is to assemble a fitting. http://www.anplumbing.com has made some good videos on this:



here they are cutting the hose with a chisel. I have not tried this method yet--I've been using a cutting wheel.

10/17/2011 2:36:26 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

here is a diagram (just a Google image search) showing positive and negative camber in a suspension.



Negative camber will have the wheel angled toward the center of the vehicle. Positive camber will have the wheel angled away from the center of the vehicle. A vehicle will have a static camber setting for when it is at-rest. When turning and/or encountering a bump the camber will change. The nature of this camber change is directly related to how the suspension geometry is designed.

Here are some charts comparing suspension behavior between a strut-type suspension and a multi-link design (both of which are independent suspension). These suspension architectures can be implemented in different ways, and unless somebody is curious I won't get into that at this point. Specifically, the strut-type suspension mentioned here is on a 1st generation Eclipse/Talon/Laser/Galant and the multi-link is on a 2nd generation vehicle.



Generally speaking, when the suspension encounters a bump on one side you want to minimize the camber changes in order to maintain stability of the vehicle. The left diagram shows how a multi-link suspension results in less camber change as the vehicle encounters bumps and uneven pavement.

During cornering, it can be advantageous for the outside wheel to experience toe-in while the inside wheel experiences toe-out. This increases the tire contact with the road. A multi-link and double wishbone (sort of a cheaper version of multi-link, but the terms get thrown around a lot) suspension can achieve this over a strut-type suspension. The strut suspension has a tendency to generate positive camber during cornering.

10/19/2011 10:50:58 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

On turbo projects a common question asked is, “Which electronic boost controller should I get?” A lot of that comes down to price and preference. Believe it or not, aftermarket external electronic boost controllers tend to be pretty similar across makes and models. The components are similar, the control systems are similar, and the process of tuning the controller is also similar. Here I am presenting a comparison chart between the most common aftermarket electronic boost controllers currently available. To illustrate how similar many electronic boost controllers really are, the chart shows the different names each model uses for basically the same type of setting or function. For example, the AEM Tru Boost adjusts the wastegate opening pressure with the “SPr” aka “spring” setting. The Greddy Profec B Spec II adjusts the wastegate opening pressure with the SET GAIN aka “Start Boost” setting. Some controllers have functions completely missing: AEM Tru Boost doesn’t have a feedback setting, and the Blitz controllers do not have a wastegate opening pressure setting. Make sure you read the instructions of a particular boost controller before you go using it.

In this comparison I did not include major standalones such as Haltech and AEM EMS due to the complexity of setting them up. I also didn’t include the old Greddy Profec B and the older HKS EVC units (anything besides EVC-S) because they use stepper motors and that is not a common design anymore.

Basic Settings

Baseline duty cycle - this is a value usually from 0 to 99 which sets a baseline ratio of ON to OFF time for the solenoid. The boost controller rapidly cycles the solenoid. When the solenoid is ON, it is operating in a way that works to keep the wastegate shut . A higher value here corresponds to higher boost levels. You should start near 0 for this setting.

Wastegate opening pressure - This is the pressure at which the wastegate is allowed to open. If the current boost level is below this value, the gate will be stay shut. The higher you set this value, the more you will improve your spool, but at an increased risk of spiking. If this value is set too high it can lead to spikes or oscillations. You should start near 0 for this setting.

Feedback – Without getting too technical, this is a correcting feature of the boost controller which usually 1) improves initial spool and 2) reduces the amount that the boost falls off at high rpms . If this value is set too high it can lead to spikes or oscillations. Feedback is NOT the same as “self learning.” Self-learning features basically attempt to do the work of tuning the controller for you. Feedback is actually another parameter for you to tune and should not be confused with self-learning features. You should start by setting this value at or near 0.

Overboost protection – This should not be confused with an overboost warning. Many controllers will illuminate or sound an audible alarm, but the alarm itself just alerts the driver of a problem. The overboost protection could be a completely separate or a separate but related function to the overboost warning.

On an external EBC the overboost protection function of the controller will disable the solenoid once a certain boost has been exceeded. The controller disables the solenoid in an attempt to bring boost down to the wastegate spring pressure. Each controller implements the overboost protection feature in its own way so be sure to read the instructions carefully. This disabling of the solenoid will not protect your engine in the event of a serious mechanical failure such as a stuck wastegate or a melted wastegate hose. On factory ECU’s or standalone engine management overboost protection is accomplished usually with a fuel cut. Unlike a solenoid disable, a fuel cut will protect your engine in the event of a serious mechanical failure in the boost control system.

Temporary boost increase – usually this is called Scramble Boost. It’s a feature that allows you to raise the boost for some number of seconds. It is implemented differently across manufacturers. This is the kind of feature you would expect Paul Walker to use when he runs out of NOS. In reality not many people set this up.

Boost target – On controllers with a Self-Learning mode, this is the intended boost pressure that the controller will try to reach using a self-tuning process. Apex’I uses this setting differently. The AVC-R can utilize a boost target setting even when Self-Learning is disabled. On the AVC-R the boost target is used as part of the regular feedback system, unlike other controllers. The AVC-R’s implementation of a boost target is closer to what you find in many standalones (AEM EMS) or programmable stock ECU’s such as a Subaru stock boost control system. The Power FC has a boost target setting but its actual real-world use is more like a rough adjustment of the feedback system.

[Edited on October 20, 2011 at 11:44 AM. Reason : .]

10/20/2011 11:42:36 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

Additional Features

Self-learning - This is NOT the same as a Feedback setting. This is a mode where the controller is supposed to figure out the settings for you, once you have picked a target boost and gone through a learning procedure. These features generally have mixed results at best. Personally I do not use self-learning modes because I feel that having someone competent adjust the controller manually will have a better final result.

Gear or Vehicle Speed based adjustment - This allows boost adjustments based on rpm or vehicle speed. Most commonly this is used as a form of traction control. These features are typically found on standalones but only a couple external electronic boost controllers have this kind of capability.

RPM-based adjustment - This allows fine tuning of duty cycle based on RPM. This could be used to keep boost steady at higher engine speeds.

After all that explanation here is the chart:



Note that some controllers combine two functions into one setting. For example, the Blitz controllers use the “Set” value as a boost target when in self-learning mode, but when in manual adjust mode the “Set” is a baseline duty cycle value.

Adjusting your boost controller

As a general strategy, when setting an electronic boost controller you should usually start with all the main settings (baseline duty, wastegate opening pressure, feedback) near 0. Configure your overboost protection setting. Now begin raising boost by alternating between increasing the baseline duty cycle and increasing wastegate opening pressure in small increments. Once you notice that the engine is slightly below the intended boost level, start adding in feedback until the target boost level is achieved and boost is relatively steady. From there you can make minor tweaks to the settings again. Every car is different but usually spikes or oscillations can be fixed by turning down the wastegate opening pressure or feedback settings.


In conclusion, the major external electronic boost controllers available today have the same basic system design. Some models have more or less features than their competitors, but all of them will get the job done as long as they are installed properly, tuned correctly, and used on a turbo system that is mechanically sound.

[Edited on October 20, 2011 at 11:45 AM. Reason : .]

10/20/2011 11:43:34 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

Sooo I'm getting a little bored with just posting stuff up here. Anybody want to suggest a topic? Ask a question? Make some comments? Basically anything is fair game...

10/23/2011 1:25:27 AM

Hiro
All American
4673 Posts
user info
edit post

Suspension setups.... how about discussing more in depth about 5 link suspensions, determining optimal ratios of f/r spring rates, shock dampening/rebound, etc..

Perhaps discussions on dry vs wet differentials?





[Edited on October 23, 2011 at 1:43 AM. Reason : .]

10/23/2011 1:39:38 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

Quote :
"Suspension setups.... how about discussing more in depth about 5 link suspensions, determining optimal ratios of f/r spring rates, shock dampening/rebound, etc.."


There's literally two dozen equations needed to determine spring rates, in the sense that one equation needs a bunch of others to calculate out all the variables. I'm looking at them right now, the "dual mass oscillator model." I've got a few charts and diagrams here which might be helpful, I could post those.

Multi-link suspension is interesting in that there are so many implementations of it. I can post some... I have the most documentation on Rx-7's and Rx-8's, which are among the most advanced and well-tuned suspensions produced for a reasonably-priced car. As I explained in an above post about the 2nd generation Eclipse, the key is that the positioning of the suspension links and the compliance of the bushings affect camber and toe changes in different driving conditions.

These days though with driving dynamics systems such as active suspension and stability control, it's cheaper for the OEM's to use electronic aids and simpler suspension design to achieve almost the same effect as expensive suspension tuning.

I've got some diagrams of monotube vs twin tube shocks. I could post that.

Quote :
"Perhaps discussions on dry vs wet differentials?"


Could you be more specific? I discussed the clutch type, viscous type, and helical/torsen type. Another design family uses a planetary gearset to distribute torque between two sides, like a torque vectoring differential.

10/23/2011 12:36:24 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Below are diagrams of the five-link suspension in a Hyundai Genesis sedan. The coupe has cheaper strut-type suspension in the front but the rear is the same.





There are a number of advantages to the arrangement used here:

1) improved ride quality

2) improved stability in cornering and braking

3) improved turn radius

4) improved interior space

The arrangement of the links creates advantageous dynamic changes in the camber, caster, and toe according to the driving condition.

10/30/2011 11:25:30 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

I can't wait till you find some info on the subie/yota direct injection/ port injection combination system

11/7/2011 2:57:16 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

I have plenty on it... what do you want to know?

just ordered a collection of papers on Corvettes. It's on sale for an insanely low price for like $18 http://books.sae.org/book-pt-118 lots of shit on LS1's and Z06-specific stuff

11/7/2011 2:59:41 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

Does it really staging the direct injection up to x and then controlling the port fuel injection after that.

i wonder how tunable it will be?

Who makes the ECU subie or yota?

11/7/2011 3:09:04 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

I have information on the Toyota D-4S system in the 2GR-FSE application, which is the engine used in the Lexus IS350. I'm sure it will be a while before any somewhat public info comes out on the new boxer engine.

Quote :
"Does it really staging the direct injection up to x and then controlling the port fuel injection after that. "


Only under low rpm and high load. Even then it blends the two. Only the direct injectors fire at high rpm on 2GR-FSE. The reason for this is that firing the port injectors under those conditions result in high temperatures at the tips of the direct injectiors. If you put bigger port injectors in you would have to change the injector staging. And then you will shorten the life of the direct injectors due to carbon buildup. Now, would it be a noticeable shortage or something that would only come up under some obscure conditions? I don't know. Remember my earlier posts about carbon buildup?

This system doesn't work the way people think it does. It's counter-intuitive in some ways, unless the system has changed significantly in the BRZ engine compared to the 2GR-FSE.

Quote :
"i wonder how tunable it will be? "


I predict that, for the first few years at least it will be either

1) a disaster.
2) a big disappointment/meh

On the disaster front: this is a complicated system and with the hype for this engine, there is chance people will rush some aftermarket tune out the door. Most likely those tunes will result in accelerated carbon buildup around the direct injectors. The driveability aspect could also be a big deal depending what people start fiddling with. GDI engines are highly sensitive to injecting timing because the mixture forms in a narrow time window compared to port injection. Most tuners don't know much about injection timing, and even if they did it's very hard to tune it without a gazillion dollar research lab.

On the disappointment front: this system is designed for a high specific-output naturally aspirated application. I very much doubt you will be able to squeeze much power out of this BRZ engine in n/a form except maybe from changing/removing the cats. The heads for example in the 2GR-FSE are specifically engineered for high flow in a naturally aspirated application. Toyota went way out of their way to not have tumble or swirl generation valves so as to not restrict the intake ports.

As for putting a turbo on it: fuel system will be a big limitation for the aftermarket unless there is major headroom in the stock system. Good luck getting bigger direct injectors. Toyota uses a different design on this system than anything out there that I have seen. It is not the multi-hole type (Ford, Hyundai), it is not the swirl type (lots of first generation engines), it is not the piezoelectric/spray guided type (BMW). It has a fan spray pattern.

On a PFI engine, injectors are for the most part interchangeable. There are only a small number of variations in use on production engines, and usually the injector sprays on to the back of a closed intake valve. So you can can get away with a lot. On a GDI concept there's less room for error.

Now I say all this but in 5 or 10 years somebody somewhere will work something out and get through most of the issues as parts and knowledge becomes more available. I would wait until an actual turbo engine cames out though. In that situation guys will figure out a few basic tables to modify in order to make the thing go faster by pushing the fuel pump and injectors harder. Don't hold your breath for aftermarket injectors for this system.

Quote :
"Who makes the ECU subie or yota?"


Highly likely it is Denso ECU hardware-wise just like so many other Toyota and Subaru hardware. Denso is like the Delphi of Japan in some ways. It's the biggest supplier for engine controls and it's actually a spinoff of Toyota. The calibration was probably done mostly by Toyota and Yamaha engineers considering that they were the ones who did the 2GR-FSE.

11/7/2011 3:37:01 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

Cool, so it's likely people will get into the ECU asap.... whether or not they know what to do with it is a completely different story.

11/7/2011 4:07:57 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

In my mind, the gap between the designs of the OEM's newest vehicles and the public's general knowledge is increasing. I think you can directly link that to increasingly tighter emissions and safety regulation. As regulations tighten at a faster pace, technology becomes adopted faster and you have to work harder to keep up.

11/7/2011 4:18:46 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

You were asking about injector staging...



x axis is rpm, y axis is load, and the chart shows the % of the total fuel delivered by the direct injectors. So 100% = direct injection only, 0% = port injection only . This is on the 2GR-FSE engine found on an IS350.

[Edited on November 7, 2011 at 4:32 PM. Reason : .]

11/7/2011 4:31:46 PM

arghx
Deucefest '04
7584 Posts
user info
edit post



You can see the "shell shaped" piston here

11/8/2011 9:41:11 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

that piston looks like it would have hot spots all over it.

11/9/2011 6:29:43 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

Why do you say that?

11/9/2011 8:26:48 AM

sumfoo1
soup du hier
41043 Posts
user info
edit post

I know they probably control burn more than i'm used to but in my limited knowledge of piston design (mostly for boosted engines) you try to stay away from sharp edges as much as possible because they heat up faster then the rest of the piston and become points of pre-ignition.

11/9/2011 8:35:33 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

The injection timing, piston shape, and intake port design all work together.

So here on the 2GR-FSE (the one with the port & direct injection)



You have the dual fan-sprays ("developed spray") hitting the piston bowl (top diagrams, 284 BTDC) and then forming a mostly homogenous mixture under full load.



Above is the stratified mixture formation on a 3GR-FSE during cold conditions. This engine does NOT have port injectors like the upcoming FA20 engine on the BRZ--it is the Lexus GS300 engine. On the 2GR-FSE, which uses the D-4S system with port injectors (like the FA20), there is a 65/35 split of direct injection and port injection during warm up. More direct injection stabilizes the engine output so you don't get vibration and possible misfire. Too much direct injection and the mixture will not form properly to keep emissions down. That's how they settled on 65%. These Toyota engines use a shallow piston bowl to get just enough stratification during warm up. If the piston bowl is deeper you get more stratification but lose WOT performance by getting rich spots near the bowl.



That's regarding the 3GR-FSE, the engine which does not have port injectors. The Ecoboost engines use a piston bowl of a little bit different design than the Toyota engines:

11/9/2011 9:26:05 AM

ghost613
Veteran
324 Posts
user info
edit post

Could I make a request?
What goes in to deciding firing order? (hopefully this isnt stupidly simple or I haven't missed it somewhere)
If you want to focus your response (if you want to respond at all) the reason I ask as that someone mentioned that the firing order for the mustang 5.0 is better than the firing order for most other fords (1-3-7-2-6-5-4-8 vs. 1-5-4-2-6-3-7-8). I guess I would assume that balance and heat dissipation, are two main factors but then again thats probably why im not an engineer

12/16/2011 11:57:38 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

no, you nailed it.

managing heat, balancing the crank, and over all NVH of the engine are usually what pics the firing order.

12/17/2011 9:15:36 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

quote from GM's paper on the LS1, talking about firing order changes from previous small blocks:

"The firing order was revised from a traditional 1-8-4-3-6-5-7-2 to 1-8-7-2-6-5-4-3 for the LS1 to reduce crank arm stresses by seven percent and to improve main bearing performance. The new firing order increases the minimum oil film thickness by 13%."

on inline engines the firing orders are often very similar across different designs

12/18/2011 6:03:27 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Time to talk about the GM turbo methanol injection system.

Engine Overview

A few of you might be aware that GM went through a rather experimental phase in the early 1960s where the various divisions produced vehicles that we normally would not expect out of the company. One was the Chevrolet Corvair, a boxer-powered rear engine car that had an available turbocharged model. Another car, the one I am going to discuss here, was the Oldsmobile F-85 Jetfire, powered by a 215 gross horsepower turbocharged high compression all-aluminum 215cid (3.55 liter) V8 using preturbo methanol injection.






The engine had a 10.25:1 compression ratio and ran about 5psi peak boost. Above you can see that the turbo was mounted above the engine and the compressor outlet bolted directly to the intake manifold without an intercooler. The side-draft Rochester carburetor and methanol injection system were all integrated into a pre-turbo (draw-through) configuration:



Note that this system did not inject pure methanol: it was "turbo rocket fluid" which contained water, methanol, and some kind of corrosion inhibitor.

[Edited on December 18, 2011 at 6:25 PM. Reason : .]

12/18/2011 6:25:22 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Methanol Injection Control



Here's how the methanol is controlled. It's kind of ingenious how they managed to do everything mechanically. As you would expect of a pneumatic methanol delivery system, boost pressure flows from the carburetor ("boost pressure signal line") to the methanol tank in order provide the force needed to deliver the methanol. Methanol flows out of the bottom of the tank, through a filter and then to the "turbo rocket fluid control valve assembly." A depressurization valve and check valve keep methanol from flowing into the engine when it is not running. The control valve assembly accomplishes all methanol metering and delivery using completely mechanical means. To begin the process, valve #2 at the bottom lifts off its seat when boost pressure reaches 1psi. When valve #2 opens it allows methanol to flow to flow to valve #3. Valve #3 meters the methanol to the engine based roughly on throttle position.

Valve #3 is connected to ported vacuum from the throttle plate area. Ported vacuum predictably increases as the throttle opens. It is typically used in some emissions control systems and sometimes spark advance systems on old style distributors. So as the throttle opens, more vacuum acts on Valve #3 and this in turn increases the amount of methanol delivered to the engine. No solenoids, rpm switches, or anything electrical is used here to control the delivery.

Methanol Failsafe Systems

The Jetfire had a system to safeguard against depletion of methanol and safeguard against an overboost condition. This safeguard consisted of a butterfly valve in the intake system called the limiter valve, and a diaphragm that could pneumatically activate this valve under certain conditions. The methanol tank had a float inside. As the methanol was consumed, the float dropped and would initially activate a warning light to alert the driver. If methanol was completely consumed, the float would drop far enough that it would result in boost pressure being able to flow through the control valve assembly to the limiter diaphragm. This butterfly valve would then cause a restriction on the inlet of the turbo to protect the engine.

The turbo had a wastegate that ran at spring pressure and did not utilize what we would now consider a boost controller. In the event of boost somehow exceeding the wastegate spring pressure, a pop-off valve in the methanol tank would open to drop the methanol tank down to atmospheric pressure. Due to the design of the system this somehow ends up activating the limiter valve which creates a restriction in front of the turbo. So the limiter valve can protect the engine in the event of an overboost, a lack of methanol in the tank, or some kind of other pressure loss in the methanol delivery system.

This cheesy 60s commercial reveals the excitement of turbo rocket power (check out that sweet boost gauge at 0:58):



With the Jetfire GM leaped well ahead of its time and became a pioneer in turbocharging gasoline engines for passenger cars. Supposedly the car itself was a commercial disaster and nobody knew what the hell to do with a methanol injected v8 family-ish car that ran on premium gas. Given the limited technology they had available at the time the system seems well-engineered with its failsafe system and progressive delivery.

Source: Lewis, "The Oldsmobile F-85 Jetfire Turbo Rocket Engine," 1962

12/18/2011 6:27:49 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

yeah i'm still debating on going e-85 or putting water/meth injection on the roo.

e-85 is pretty easy, just requires big injectors, fuel pump, etc.
where water/meth requires a tank, nozzle, low limit switch, pump etc.

And when you run out of e85 and nothing tells you, your engine stops...

when your run out of water/meth and nothing tells you, your engine blows up.

12/19/2011 11:35:03 AM

arghx
Deucefest '04
7584 Posts
user info
edit post



Above are the designs Chrysler used on the legendary 426 Hemi engine in the mid 1960s. The lower left was the original header design used for racing applications including the 1964 Daytona 500. It had 40" runners measured from the exhaust valve to the 4" collector. The above left design later replaced this for racing applications. It had 30" runners collected closer to the engine, which apparently improved output on a dyno.

The design on the right is the cast iron manifold used on the 426 sold for street applications.

12/28/2011 4:47:50 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

We were talking about firing orders before. The Ford 5.0, BMW 4.4 twin turbo, and Porsche 4.8 twin turbo all have a firing order of the 1-5-4-8-6-3-7-2. All three of these engines have a 90 degree angle between the cylinder banks. The firing order is the same as the old flat head Fords. I presume it is for balance. This firing order is kind of weird and it presents some problems for exhaust manifold/turbo manifold design.



In the diagram above the solid lines show how the firing order progresses. The dotted lines show a major source of exhaust backpressure--the blowdown exhaust pulse. During the power/expansion stroke, as the piston nears the bottom of the cylinder the exhaust valve begins to open (the blowdown event). Opening the valve releases high pressure exhaust from the cylinder and these pulses can flow back into the previous cylinder in the firing order. That in turn causes all sorts of problems, especially in a twin turbo application with conventional exhaust manifolds.

On a normal inline 4 cylinder each cylinder fires 180 degrees apart and exhaust backpressure can be minimized through careful control over valve timing and exhaust manifold/turbo manifold design. On these V8 engines cylinders #2, #8, #1, and #6 are disproportionally affected by high exhaust backpressure. This makes valve timing (cam design) and turbo manifold design difficult.



The chart above shows exhaust port pressure on the right bank of these engines. 90 degrees of crank rotation after cylinder #2 fires, it must work against the blowdown pulse of cylinder #1. Cylinder #1 must then work against the regular exhaust flow (not the blowdown) of cylinder #2 while #2 is in the exhaust stroke. Cylinder #1 must also work against the blowdown pulse of cylinder #4. In contrast cylinder #4 does not have to work against a blowdown pulse at all because it is 270 degrees of crank rotation apart from cylinder #3.

BMW has addressed this issue by putting the exhaust manifolds inside the V and using two twin scroll turbos. Twin scroll turbos have divided inlets and are fed from two different sources, either two cylinders or two groups of cylinders. The exhaust manifolds on the BMW 4.4 are designed such that each cylinder is completely insulated from blowdown pulses. This design is complicated and expensive to implement however.

Porsche uses exhaust manifolds in the conventional location but has special camshafts to change the valve timing. Without going into too much detail, this helps backpressure and interference for most of the cylinders but #2 and #8 are still significantly affected.

1/1/2012 9:27:51 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

awesome intel i love reading these things

1/1/2012 10:40:47 PM

moron
All American
33692 Posts
user info
edit post



Do they confirm these computations experimentally?

1/2/2012 1:02:58 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

^ they have transparent single cylinder engines and other equipment used to verify the spray patterns. Here is a spray pattern for the BMW N54 direct injected engine:



the fluid dynamics software allows them to reduce the man hours and the number of actual prototypes that need to be built before settling on a final design. There is a whole methodology used called fractional factorial design experiments http://en.wikipedia.org/wiki/Fractional_factorial_design

1/2/2012 10:04:15 AM

sumfoo1
soup du hier
41043 Posts
user info
edit post

my thought is that lip, after repeated cycles is going to be a hot spot in the piston.

1/2/2012 10:13:55 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

If enough deposits build up in any particular area then yeah there could be a hot spot developing. That's why there needs to be good oil separation in the crankcase ventilation system, good spray targeting inside the cylinder, and various methods to keep the piston from getting too hot. One of the purposes of a piston bowl is to actually give more space for the flame to propagate, instead of the flame colliding with the piston and quenching.



This is chart compares piston designs between Mazda's first generation and new Skyactive direct injection systems. Note that both systems use a different type of injector than what we have been discussing. The engine is a 2.0 MZR non turbo used for research purposes. You can see that with the old convex design, the flame quenches on the piston top. With the piston bowl the flame continues burning.



The piston bowl design has a better overall heat release rate and faster combustion.

1/2/2012 10:34:35 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

The chart below was developed by GM during its development of the 6L80E automatic transmission as currently used in its V8 applications. It shows the relationship between a component/subsystem's length/radius and its torque capacity.



A higher number expresses a greater sensitivity for that component in an exponential sense. A "1" indicates that torque capacity would increase linearly with an increase in a given dimension (radius of input shaft for example). A "3" would indicate a cubic relationship. You can see that the input/output shafts, differential (on FWD/AWD applications), torque converter, wet clutch packs, and conventional one-way clutches are especially sensitive to sizing. This makes automatic transmission design tricky due to weight and packaging constraints, especially when trying to create a transmission family that can used in variety of applications.

[Edited on January 2, 2012 at 10:38 PM. Reason : .]

1/2/2012 10:36:51 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Let's talk about the VTEC system of the Honda R18A1 engine which is found in the 2006+ non-Si Civic. This is the successor to the D series engines found in earlier Civics. The VTEC-equipped D series used the low and high speed cam configuration like the NSX (see first post).

The R18A1 uses a VTEC system designed specifically for fuel economy by controlling one out of the two intake valves per cylinder. Under "high output" operation both have mostly conventional valve timing and lift--it's a "normal" DOHC inline-4 engine. For improved fuel economy, the engine has a "delayed closure" cam.



The delayed closure cam is an inexpensive implementation of

1) increased charge motion (swirl effect)
2) Miller/Atkinson cycle processes
3) Reduced pumping losses by operating mostly unthrottled

You can see in the above diagram that one of the intake valves has the same valve opening timing but with increased valve lift, longer duration, and a later closing timing. Greater lift and duration induces a swirl effect (like stirring coffee) similar to what is used on systems like Subaru's AVLS:



During delayed closure operation, one of each cylinder's intake valves stay open through part of the compression stroke. This blows part of the mixture back out of the intake port, limiting engine output and decreasing the compression ratio. The key here is that the compression ratio decreases but the expansion ratio (during the power stroke) mostly stays the same, which ultimately improves engine efficiency. This process is used in Mazda's Miller cycle engine (Millenia), Toyota's engines for hybrids (Prius) and now Nissan's VK56VD direct injected engine for the Infiniti M56.

The late intake valve closing works in conjunction with the electronic system to reduce pumping losses. On a conventional engine the electronic throttle controls engine torque output by creating a restriction on the intake (which also produces intake manifold vacuum). The engine then has to do additional work to overcome the restriction. An electronic throttle system can open the throttle valve wide in order to reduce this restriction, but then something must be done to reduce engine output for lower load driving. Late intake valve is one method of doing this, and the R18 uses its VTEC system to accomplish this. Other methods include continuously variable valve lift (Nissan, BMW, Fiat) and using the fuel injection duration and timing (stratified charge direct injection which have some similarities to diesels).

During the development of the R18A1 engine, Honda studied the effect of various intake valve closing timing for that intake valve per cylinder.



The gray line on the left part of the chart is the conventional intake valve closing timing used for the higher output VTEC mode. Then as you look along the x axis of the chart you see increasingly late intake valve closing and its effects on a bunch of different factors. In all 4 plots, a lower value on the Y axis should be considered more desirable. The top is Pumping Mean Effective Pressure--this represents work done inside the cylinder wasted on pumping air in and out. The second is brake specific fuel consumption (grams of fuel per kWh of power). The third is coefficient of variation of indicated mean effective pressure, a measure of combustion instability. The last is combustion phasing--the crank angle at which a certain percent of the in-cylinder mass is burnt.

Look at the gray line on the right and you will see why the intake valve closing was set the way it is. Remember that the intake valve OPENING is the same. Honda retarded the intake valve closing up to the point where combustion started to significantly deteriorate.

This work is called pumping mean effective pressure (PMEP) and it is a metric for determining the amount of work wasted on pumping air in and out of the cylinder.

1/7/2012 2:01:21 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

any intel on the next m3 electric spooled turbo/generator setup ?

1/9/2012 1:51:15 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

I've found some stuff on a very recent GM study for using that kind of system and I'd like to dig into it at some point. I also have a bunch of Honda AWD system stuff I want to look at, and some older Cummins diesel info.

I've been looking over cylinder deactivation systems and direct injection spray patterns lately. On the subject of the latter (and I've covered parts of this before), here are the major types of direct injection spray patterns in use today:

Types of GDI Injectors

Here are the major types of GDI injectors in use. The first and oldest type is the swirl-type solenoid injector:






It has a cone-shaped spray pattern and is side-mounted in the combustion chamber. The early Toyota D-4 direct injection system used a swirl injector as well as a lot of other 1st generation GDI designs. I believe the Mazda MZR engine in the Speed3/Speed6/Cx-7 uses this design. Manufacturers have mostly abandoned this design on the latest generation of mass produced GDI engines. The newer injector designs promote better mixing in the combustion chamber.

Next up is the center-mounted (next to spark plug) BMW piezoelectric injector:





This type of combustion system uses a more expensive, higher pressure, and more responsive injector. It does not depend on charge motion (swirl, tumble) for mixture formation. The next type of injector, the multi-hole solenoid type is by far the most common on the latest generation of GDI engines:







Note that the above two images show a 6-hole injector. Many variations of multi-hole injectors are used in production engines, with different numbers of holes and spray patterns. Ford, Hyundai, GM, VW, and others use some kind of multi-hole injector on their current GDI engines. Finally we have the fan-type solenoid injectors used in all but the earliest Toyota GDI systems. Here is the fan type injector with a single spray:





Finally, here is the dual fan-spray type used on the 2GR-FSE Lexus IS350 engine, which uses the most recent Toyota GDI system called D-4S:



You can see that the injectors used in the D-4S 2GR-FSE engines are significantly different from basically everything else out there right now. The 4 cylinder boxer used in the new Subaru BRZ/Scion FR-S will likely use this style of direct injector.

1/9/2012 9:35:25 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

The diagram below shows firing patterns on a Honda J35Z2 engine, which is the V6 engine available in the 08+ Accord



You can see that this engine can either disable an entire bank of cylinder or just one cylinder on each bank. By "disable" I mean that the fuel cuts off and the valves do not open. Most cylinder deactivation systems, including older Honda designs, only disable an entire bank of cylinders. Chrysler and GM also have cylinder deactivation but it's on their pushrod engines and it disables an entire bank. The graph below illustrates the pumping work performed during 6 cylinder & 3 cylinder modes for the Honda J35Z2:



This diagram shows that a lot less work can be performed per cylinder in the fuel saving modes. The left diagram shows the deactivated cylinders. Very little work is being performed (tiny area under the curve). The middle graph illustrates work being performed in an active cylinder during 3-cylinder mode. Smaller area under the curve, especially the bottom curve, shows reduced pumping loss. The diagram below illustrates the control strategy in action:



One of the biggest issues with cylinder deactivation is keeping the engine running smoothly and making the system imperceptible to the driver. Looking left-to-right across the graph you can see countermeasures taking effect as the cylinders switch on and off. During the switching process the ignition timing retards and the transmission torque converter lockup clutch releases. Timing retard is common during normal automatic transmission shifts and it may also be common to release the lockup on the newest generation of automatics. You can also see that during cylinder activation/deactivation the electronic throttle also opens, probably to smooth torque output and possible reduce pumping losses.

1/10/2012 8:29:31 PM

Quinn
All American
16417 Posts
user info
edit post

Quote :
"By "disable" I mean that the fuel cuts off and the valves do not open. "


This sounds horrible.

1/10/2012 9:43:53 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

no different than a Z06, SRT-8, etc. The Hondas are overhead cam so the hydraulic circuit and valvetrain is way different but it's the same final result.

1/11/2012 7:30:19 PM

arghx
Deucefest '04
7584 Posts
user info
edit post



The [cropped] diagram above shows the basic intake manifold and throttlebody arrangement for the Toyota 1LR-GUE V10 engine used in the Lexus LFA. It is a V10 engine with 10 individual electronically controlled throttlebodies and conventional port injection that has been optimized to allow 12:1 compression ratio on normal premium fuel (91 USA octane). The throttlebodies are located after the intake manifold plenum, near the intake ports. BMW uses a similar arrangement on some of its engines.

1/15/2012 6:39:15 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Today I will compare the overall hybrid system on the Hyundai Sonata to two other well-known hybrid/electricish architectures: the basic Prius system (which has been licensed to other manufacturers like Ford and Nissan) and the Volt system. The Hyundai hybrid system has a very different overall system concept especially because it uses a conventional automatic transmission.



On the Sonata you have a conventional gasoline internal combustion engine ("ICE"), an electric motor ("Motor"), an automatic transmission sans torque converter, and a belt-driven generator ("ISG"). The electric motor and automatic transmission are directly connected. An engine accessory belt drives the generator. The engine and electric motor are connected through a computer-controlled clutch. A complicated control system determines how and when the clutch engages, including clutch slip (speed difference and torque transfer between the two systems). All this is very different from the Prius-style system:



The generator (also called "MG1"), the electric motor ("MG2"), and engine are all connected together through the Power Split Device, a planetary gear set. The sun gear connects to the generator, the ring gear connects to the electric motor, and the planetary carrier connects to the gasoline engine.



There are a bunch of different combinations of how the gearset can operate to accomplish electric-only driving, combined electric & engine power, or regenerative braking. Compared to the Sonata the Prius-style system does not use a belt to drive the generator, and the Prius has a significantly more powerful generator than the Sonata. The Prius also cannot run in electric-only mode at the same speeds as the Sonata; the Sonata can run on electric power up through about 60mph. The Sonata system can be considered a simpler and less expensive system than the Prius.

In the Volt thread I described in detail how the power split worked (message_topic.aspx?topic=528749&page=10 page 10).



The Volt has a planetary gearset, two clutch packs, and one holding clutch pack to keep the ring gear from moving. The Volt can directly connect the engine to the generator, while the Sonata can directly connect the gasoline engine to the electric motor. The Volt was designed for much greater electric-only range and has the whole system built around that concept.

1/20/2012 6:59:18 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

Below is a diagram of the Lexus LFA drivetrain:



The LFA has the transmission/transaxle in the rear, a design that has been used in the Corvette since the C5 generation as well as other cars. The LFA uses an Automated Manual Transmission (AMT) made by Aisin. This type of transmission is like a conventional manual in the sense that it has a single dry clutch, but the shifting and clutch pedal operation are computer-controlled. The transmission concept has been used by BMW and Lamborghini for example. Here is a diagram of the AMT hydraulic control system:



The control system consists of three parts:

1) HPU - the Hydraulic Power unit contrains a pump and accumulator as the main supply of pressurized fluid. The clutch solenoid valve regulates hydraulic pressure to engage and disengage the clutch disc. The master solenoid valve regulates hydraulic pressure used to shift gears. A sensor informs the control unit of pressure in the master solenoid valve circuit.

2) CSC - the Concentric Slave Cylinder physically releases and engages the clutch using a piston. A sensor informs the control unit of pressure in the slave cylinder.

3) GSA - the Gear Shift Actuator replaces a normal gear shifter found in a conventional 6 speed manual transmission. It has two pistons, two pressure sensors, and two solenoid valves. One set which includes the Shift Piston moves the selector forward and backward (say from neutral to 3rd or 4th gear). The other set includes the Select Piston to move the gear shifter left and right.

A sophisticated control system regulates pressure in these various solenoids and hydraulic circuits to balance shift speed and response with smoothness of engagement, based on the mode selected by the driver. The AMT control system communicates with the engine control module so that engine torque can be increased or decreased as shifts occur.

1/22/2012 9:12:35 PM

sumfoo1
soup du hier
41043 Posts
user info
edit post

they used that transmission because the dsg was too smooth and didn't feel right from what i've read.


seems like the dual clutch setup would be better for an engine that winds that high that fast, but maybe that was the problem... the dual clutches became too much of a flywheel for the high-strung rev happy engine

1/23/2012 8:54:59 AM

arghx
Deucefest '04
7584 Posts
user info
edit post

I'm not sure... I imagine it's possible to have a harder, more sporty dual-clutch by playing around with the calibration of the transmission controller. I suspect the single-clutch design was lighter overall and there is some mention in the literature of the low inertia for the specially designed clutch disc and pressure plate

1/23/2012 7:16:36 PM

arghx
Deucefest '04
7584 Posts
user info
edit post

So VW is releasing, if they haven't already, a new 1.8 (EA888) TFSI engine with some significant new features. I'm not 100% sure if all of these features are new to a production VW engine. The engine has electronically controlled piston oil squirters that can be disabled for improved warmup. It also has a variable cam phasers and two-stage variable valve lift on the exhaust cam. Adjusting valve lift on the exhaust cam is a little unusual. The cooling system is a compromise between a traditional mechanical water pump governed by engine speed and a fully electric water pump that you would find on a BMW for example:



water temperature is adjusted by changing the position of rotary valve #1. In this way the water temperature can increase under light loads in order to reduce internal friction for improved fuel efficiency.



The electronically controlled piston cooling jets and water temperature control system improve warmup during emissions and fuel economy drive cycles such as the New European Drive Cycle (NEDC, red line below):



The new engine has some significant features for the turbocharger: an electric wastegate actuator, electric air bypass/diverter valve, and a milled compressor wheel:



Now, check out the water-cooled divided turbo manifold:





The water-cooled turbo manifold allows the engine to run leaner under heavy loads, meaning lambda=1 or 14.7:1 air-fuel-ratio on pure gasoline under almost all warmed-up conditions. Now here's the real kicker... I saved the best for last...



Volkswagen has adopted combined port injection and direct injection, just like Toyota. This allows more flexibility with mixture formation under part-load while having the benefit of being able to spray fuel on the back of the valves to limit buildup of deposits. Previous generation Volkswagen direct injected engines have a reputation for carbon buildup and misfire problems, so this may go a long way if the system works correctly and is adopted across most engine families.

Based on the document I have there are three main differences I can see between this system and Toyota's D-4S direct injection system. First, the injector is of the multi-hole type likely sourced from Bosch. Toyota uses the dual-spray fan-type injector. The second difference is in the injection strategy during warmup. On the Toyota system, the port injectors and direct injectors fire simultaneously (at least they do on the IS350 engine). On the VW system, the direct injectors fire three times per 720 degree cycle (like a direct-injected BMW) and do not utilize the port injectors. Finally, VW uses a tumble control valve to increase charge motion under certain conditions while Toyota does not.

In typical VW fashion, the horsepower output for such a complicated engine is of course laughable although it builds torque quickly.



for those too lazy to google a unit converter, that's 167hp and 236 lb/ft which is something you'd expect out of an 80s V8. It might respond well to mods.

[Edited on January 25, 2012 at 8:46 PM. Reason : dyno]

[Edited on January 25, 2012 at 9:04 PM. Reason : pic of turbo]

1/25/2012 8:38:43 PM

Hiro
All American
4673 Posts
user info
edit post

Quote :
"that's 167hp and 236 lb/ft which is something you'd expect out of an 80s V8."


I was reminded of late 80 (third gen) Camaro RS when you mentioned that.

1/25/2012 8:58:10 PM

 Message Boards » The Garage » arghx's technical ramblings thread Page 1 2 [3] 4 5 6 7 8 9, Prev Next  
go to top | |
Admin Options : move topic | lock topic

© 2024 by The Wolf Web - All Rights Reserved.
The material located at this site is not endorsed, sponsored or provided by or on behalf of North Carolina State University.
Powered by CrazyWeb v2.38 - our disclaimer.