sweeet!

 

The 16 to 17 percent factor I mentioned is a generally accepted estimate of drivetrain losses(from crankshaft to wheels) in a 2WD. I think 20 percent is used for a 4WD. They have nothing to do with the dyno.

As others have mentioned, it is not an exact science. The baseline power of the car can also vary.

 

Todd, Will there be a seperate system for the 997.2 C4S ?

 

Mike will the pricing be the same as your current 911 systems on your website?

Looking forward to the final products.

 

The exhaust system between the 2 and the 4 S are the same.

 

Interpreting and relying on chassis dyno data is tricky business.

First, let's get one thing straight: Just like any measuring tool, a dyno's data is only as good as the operator. Flawed or casual test methods will result in invalid or meaningless data. This applies not only to the dyno protocols, but also knowing the particulars of the car being tested. For example, it takes us several dozen pulls to establish a valid result for a given state of tune, all the while monitoring fuel injection parameters to ensure engine stability.

Second, not all dynos measure power in the same way. For example, a Mustang dyno measures via an eddy current load method, similar to the way an engine dyno works (which is how an OEM determined their power claims). Yet, a Dynojet uses an inertia method that calculates power based on how quickly the known mass of their rollers are spun to a given speed. These two different ways of measuring can give very different results.

Third, please always remember that even with the best operators and the same dyno brand, it is often not possible to directly compare dyno to dyno. Even with the same "operating system" there are certain required and optional user inputted values. These inputted values can greatly affect results.

We are acutely aware how inconsistent and unregulated use and cross brand calibration has tarnished the image of the chassis dyno. We have owned and operated multiple chassis dynos since 1998, and have worked hard to educate both the industry and the consumers. We have acted as chassis dyno cell set-up consultants to many companies. In responsible hands, a chassis dyno can be a useful and very important tool. Unfortunately, chassis dynos are typically not used by responsible parties.

For further reading, see this FAQ on our website:

http://awe-tuning.com/pages/faq/awefaq_main.cfm?FAQ=22


The bottom line: if a thorough test methodology is established and used consistently, accounting for dyno cell and vehicle variables, it is possible to compare "before and after" data from a single given chassis dyno.

Further, if the same rules are applied to another dyno, rudimentary cross dyno comparisons can be made in a percentage gain manner only. For example, if a known good stock baseline is done on a given dyno, and a known good modification test is done, the results can be expressed as a percentage by taking a given "after" dyno data point and dividing it by the same rpm "before" data point. Do the same on another dyno, and the resulting percentages can be *cautiously* compared.

 

Hi Todd, thanks for the reply. I have a couple of questions for AWE:

1-I know it is up to the dyno operator and you can be bit by correction factor anomolies if the correct weather info for example is not input correctly. Do you use a weather station on your Mustang dyno or do you do your own correcting?

2- I have been reading and hearing about X and H pipes for many years since my Mustang and Vette days back in the 80's. You mentioned that you lost power with an X pipe on a Gallardo. I always read that the cross over H or X in the exhaust system must be in the perfect place, thus tuned.
For the 2009 997S car I have heard that the center primary muffler is not a shared can and is 2 simple U bends...like 2 seperate muffler in 1. I have also heard that this muffler is a shared can and that the exhaust gasses mix like a X pipe would. Did you cut the primary center muffler apart to see what is on the inside? Maybe Porsche figured out the exact pipe length were this shared primary muffler (if it is shared) should occur hence the center primary muffler?

Your thoughts please? Also how is the develpoment going? Are you all still fitting parts? It sure was nice for the Porsche dealer to give you a car. How long do you get to keep the car to develop an exhaust for?

thanks, Mike

 

Hi Mike,

Our dyno will output both SAE corrected and uncorrected results each run. The corrected results are based on ambient values that we manually input. On a normally aspirated engine, we input temp, humidity, and barometric pressure.

We publish SAE corrected results.

Correct, and here is where the mid engine and rear engine configuration differs radically from a front engine configuration: exhaust tubing length.

As you have undoubtedly seen on domestic cars, the x-pipe or h-pipe is situated where there is very little distance between the two banks of tubing to begin with. The idea behind an x-pipe or h-pipe is to enhance the cylinder scavenging of one cylinder bank by harnessing the energy of the exiting exhaust pulse in the other bank.

Placing the cross over pipe at a place where there is not much distance between the two banks of tubing means that maximum energy is maintained to create this enhanced scavenging effect.

The problem with a mid or rear engine car is that there is no natural place where the two exhaust tubing banks are running close to each other.

Since the mid and rear engines are already very close to where the tailpipes exit vs a front engine car, it is rather un-natural to increase tubing length just so the two tubing banks can come close to each other and create a place for an x or h-pipe. That is really robbing Peter to pay Paul. You have added extra tube length and restrictive bends in an attempt to enhance the delicate balance between cylinders and increase scavenging.

Plus, the energy in the exhaust has been disrupted so much by the increased tubing length and bends that when it finally arrives at the x or h-pipe, it is behaving much differently than with a shorter run, which can now cause the separate bank pulses to collide instead of compliment each other. This can be seen on the dyno as the x-pipe becomes locally much hotter than the surrounding tubing.

We used a boresope to view the can internals. Yes, the center muffler on the 997.2 has a u-bend at each end of the can, so the actual internal tube runs are quite far apart even at the apex of these bends. The shared can acts as a cross over in the sense that it is probably improving acoustics, but not doing anything for flow. We think the can is there to dial out some harmonics that cropped up when Porsche moved the cats close to the head in order to improve cold running emissions.

We are still very active with this project. Several other prototyping jobs are being held up until this one is complete, as we have allocated all our engineering and fabrication resources to it.

Use of a vehicle for prototyping purposes varies. We'll probably have this car for over 6 weeks.

 

Thanks for the reply!

 

Updates:

I thought it would be interesting to share how we design and manufacture our jigs on which we ultimately build our products.

Years ago, we used to do our whole prototype process by "hand", meaning that we would cut and weld mockups and modify them for performance and fitment. And then once we had a final design, we would hand build an apparatus by which we could reproduce the part in volume, all the while trying to ensure the construction was as efficient as possible. It was a very laborious and lengthy process.

Since we interfaced our manufacturing with our pre-existing CAD/CAM software 3 years ago, our development process has steadily gotten faster and faster. That, along with the acquisition of more CNC equipment, really has streamlined our R&D and ultimately we are producing better products from our manufacturing line.

As I stated in my previous post, we first digitize the entire muffler and header/cat cavities of the vehicle, as well as the actual components themselves. This allows us to "virtually" work with the vehicle and products on our PC workstations, which in turn allows us to solve performance and production challenges. For example, instead of physically mocking up headers on the car, we can solve tubing runs for equal flow and manufacturing simplicity on the computer. Much faster!

Correspondingly, we can quickly design and manufacture the jigs that the parts will be produced in since we have a full 3D mockup of the exhaust already in virtual space:







Here is the pathway of our jig building:

First, we CAD (Computer Aided Design) render the jig parts based off our 3D exhaust models. Then in our CAM (Computer Aided Manufacturing) program, we instruct our CNC (Computer Numerically Controlled) equipment to actually make these parts . CAD -> CAM -> CNC.


Here is a jig part program in our CAM suite:





Here is the same part being cut by our CNC horizontal band saw based of that program:





Here is that part that was cut by the CNC saw being flycut for true flatness and being drilled and tapped for fixture interface in our CNC mill:





And here are similar jig parts assembled into a functional jig:





And then finally a prototype part that is fitted to the car:





Ultimately, we will dyno and street test these parts to ensure performance and sound quality, and if the need to redesign any component ocurrs, we can easliy do so since the whole R&D process is now modular. We can zero in on the component that needs a tweak instead of tearing the whole system apart.



If you have any questions about these processes or designs, do not hesitate to ask.

 

Hi Todd, does your software allow you to model the gas dynamics of the exhaust system so you can design for optimum pulse tuning and scavenging or is this still largely a trial and error exercise?

 

(photos deleted to save space)

I am truly impressed by the care, quality and expertise of the work of a 22 person company. You've brought me a lot closer in making the final call in choosing a new exhaust system for my 2005 CarreraS (Cab).
I actually am researching the whole turbo versus supercharger route - with associated exhaust system ramifications. Advice always welcomed.
Re the points/discussion about the challenge in dyno compares, I agree. It's a huge challenge to standardize and/or normalize the metrics around dynos. Even doing the relative compare and trying to normalize the output based on a reliable/repeatable test subject presumes linearity in the variance between the dyno subjects. Good luck So, dynos and good/repeatable science just don't go together. Oh well.
Actually it would be an issue if you wanted to collaborate with someone doing joint analysis and research work elsewhere. But that's why science and engineering are different disciplines
Sorry if this is a bit over-geeked.
Verde

 

We can model gas flow with our CFD program, but in reality, computer modeling gas and fluids is still an imprecise science, so we rely more on actual power testing when it comes to overall system performance.

Our CFD software works well when modeling individual components, like a header primary.

 

Hi Verde,

Thanks for the kind words. It is much appreciated.


The VF Engineering kit (formally marketed under the Evo name) is really the most comprehensive and best engineered forced induction kit on the market for this vehicle. Truth be told, we have not designed a specific exhaust system to work with the supercharged cars, but VF has told us that they are endorsing our products based on in house testing they have done.



Well, dynos and good science CAN go together, it is just unfortunate that so often they do not in this industry. I think that is because of a few factors.

Even though dynos are relatively expensive, they are most often purchased as just another piece of shop equipment and used as a money making tool, not as an R&D tool. The far majority of chassis dyno owners these days are not part manufacturers. They are "tuning shops" that install other companies' parts. Therefore, these shops are "once removed" from the scientific potential of these machines. And the manufacturers of these machines have realized that comprehensive scientific training is lost on these types of shops, so there is no "standardization" of the chassis dyno purchasing experience. You can be trained and learn how to use a chassis dyno as a scientific research tool, but the vast majority of users don't care about that. They just want to strap down client cars, do three power pulls, and collect the money. Such a lack of standardization is a recipe for flawed and unreliable data generation.

Regarding the variance in test subjects: yes, there can be variance among cars, further complicating cross dyno comparison, but in reality this variance seems to be very small, especially with the highly engineered German cars. We have seen this simply by dyno testing a lot of the same models here, while being consistent with our test protocols.

 

Interesting.