Independent testing of the Grimmspeed Intake

[Tuning Alliance]

How many pulls did you do per setup? Based on how your first post is worded, it sounds like you did only 3 runs total. If you did more then 3 runs (which I'm sure you did, just need to reword the first post), then on which run per setup is the power numbers shown for?

 

How much time spent between pulls?

 

Was the hood closed for all of the pulls?

 

Did you perform an A-B-A test? Where for the last couple pulls you install the original part and see how it performs? This is important since the car will usually will continue to loose power as you continue doing back to back pulls.

 

What was the ambient temperature, barometric pressure, humidity level during this test?

 

What base map was used for the stock intake? Was the timing and fuel tables left the same during all of the runs? Did you try to tune for max power for any of the setups?

 

Like I said I'm OCD, but I also want to see good accurate data before I shun a part as "useless".

 

 

As I stated in my first post the controlled settings were used to compare the intakes. So with the stock intake installed, I left the Maf curve completely stock. I loaded a set timing map and edited the primary fuel table to get a flat 10.8 AFR. So this took several pulls, starting with partial pulls to get dialed to a nominal setting. I can look up exactly how many if you think this is highly important. Once the settings were good, I did a pull to 6200rpm. There were roughly 3-5 minutes between runs. Aside from when we were swtiching intakes, between intakes it was about 15-20 minutes so the engine had decent cool down time between.

 

Then I installed the K&N panel and did a full pull it was about 11.0 AFR and made very good numbers. I added to the maf curve over another 2-3 runs to get a flat 10.8 AFR. Since it beat the stock intake power levels we decided to move to the GS intake.

 

We installed the grimmspeed intake and I added 12% ontop of what was already added for the K&N panel. Did a pull and it pulled timing. I will add more detail to my first post. Moral of the story is the GS intake could not beat the panel with its provided air box.

[Tuning Alliance]

And it all reality if you looked closely at the graphs in the first post you might notice the grimmspeed run did have additional timing. I never changed the timing map during the comparison, so I believe the difference in MAF calibration altered the calculated load enough to be running slightly more timing. I will have to verify this and display what the calculated load was.

[wicknetzel]

We installed the grimmspeed intake and I added 12% ontop of what was already added for the K&N panel. Did a pull and it pulled timing. I will add more detail to my first post. Moral of the story is the GS intake could not beat the panel with its provided air box.

 

GS stated that it was a 12% scaling from the stock intake. Didn't you say you changed parameters for the K&N panel? If so, then isn't the 12% scaling from that point different than what they talked about?

[Tuning Alliance]

GS stated that it was a 12% scaling from the stock intake. Didn't you say you changed parameters for the K&N panel? If so, then isn't the 12% scaling from that point different than what they talked about?

 

 

From my first post "Calibration of the Grimmspeed intake was found to be 12-13.5% above the stock MAF curve (12% from 1-4 volts tapering up to 13.5% @ 4.1-5 volts). "

[serx7]

This. You are comparing two completely different things when you talk about using the stock flat panel versus a round intake in terms of turbulence. When you consider the flow stack design, you want an equal amount of air being able to enter from the entire circumference of the intake. Throwing induced air into the mix would have to be in such a way to not create turbulence, and cycling air around the box would likely do this.

 

IMO, you're better off chasing after blocking the heat from the headers than trying to induce airflow. We're also talking about IATs on a dyno... which is not reflecting real road atmosphere. Not to mention, ducting only does so much at a standstill, which seems to be people's concern here.

 

Last thought in regard to the design is this: people who are ACTUALLY engineers designed this and thought very hard about it. That's what you paid for. Before you hack up your new intake... I'd talk to them about your concerns.

 

Here's what I can tell you in my case. I had a reputable tuner (popular here, but don't want to drag them in unless they volunteer) do my stage 2 setup (uppipe, downpipe, GS TMIC) and all that I did was scale my MAF by 12% from that. When I compared datalogs, I saw slightly earlier boost and a consistent 1 psi increase once I got past peak boost (my vf40 was already maxed out, so I wasn't expecting to see a higher peak boost). Keep in mind, 1 psi goes a longer way when you have a larger intercooler and are flowing a larger number of moles of air over the stock TMIC.

 

Agreed. Not planning on touching the GS airbox or the stock air duct (in case I ever decide to revert back or sell off the stock airbox assy).

 

Over the past ~week of driving my normal routes I can't say I've observed IAT being different from w/ the stock airbox in stopped/slow moving traffic so far. After constantly cruising for 20+ min above 40mph, IAT settles to within 5-8F of what the dashboard ambient temp reads, and after several minutes of driving in slow traffic, IAT doesn't seem to increase much beyond 10-12F of ambient before traffic starts moving again and IAT starts to decrease again. I realize the dash display isn't the best way to measure true ambient b/c of potential heat xfer from the engine bay, but the delta doesn't seem to be different from what I observed w/ the stock airbox and upon cold start the IAT and dash ambient displays match within 1F.

 

I'll see how things play out over time, and assuming I'm still running the GS in the spring/summer next year, I'll see how IAT behaves when ambient temps are much warmer than they are now. If anything I'd probably look into better shielding or drilling a few more drain holes in the wheel well trim extension that runs below the foglight assy.

[heiche]

Moral of the story is the GS intake could not beat the panel with its provided air box.

 

I'm surprised by this. Was expecting some additional power. Hard to argue dyno results though, even though air temp characteristics can be different on a dyno, sounds like you made a good effort to minimize those differences.

 

You should make this a series, with different parts, header, turbo, etc - seriously. Before + after dyno plots generate a lot of discussion and thought

[covertrussian]

We installed the grimmspeed intake and I added 12% ontop of what was already added for the K&N panel. Did a pull and it pulled timing. I will add more detail to my first post. Moral of the story is the GS intake could not beat the panel with its provided air box.

 

Was the timing pulled due to knock or due to the IAT comp table? Stock map for an 05 starts pulling 2* at 104F, since this is a tuned car it's probably pulling more by that temp.

 

Lack of airflow to the fenderwell on the dyno does seem to be the main issue with this, this is where doing open road logs of intakes and then plotting them into Virtual Dyno would provide more real life numbers.

 

 

And it all reality if you looked closely at the graphs in the first post you might notice the grimmspeed run did have additional timing. I never changed the timing map during the comparison, so I believe the difference in MAF calibration altered the calculated load enough to be running slightly more timing. I will have to verify this and display what the calculated load was.

 

This is why I'm thinking the intake made the engine go past MBT. Your test showed us is that the new intake didn't do well on someone's custom timing map which seems to be fine tuned for another setup.

 

I believe a much more accurate test would have included fine tuning each setup and seeing the power restrictions of each intake.

 

I know I sound like a GS rep here, but I didn't even buy their intake (for my own reasons), in fact I only own 1 product from them (crank pulley). The reason I'm questioning your testing methods is because I saw similar gains to GS with my custom intake, which told me that stock airbox, while a great cold air intake, is not a great flowing intake.

 

GS stated that it was a 12% scaling from the stock intake. Didn't you say you changed parameters for the K&N panel? If so, then isn't the 12% scaling from that point different than what they talked about?

 

They are running close to 14% from the sounds of it, which would put in a different column/cell. But you shouldn't treat the 12% as an absolute number, you add MAF scale until you are AFR's are close to what they were before, which for some cars might be way more then 12%.

[serx7]

GS stated that it was a 12% scaling from the stock intake. Didn't you say you changed parameters for the K&N panel? If so, then isn't the 12% scaling from that point different than what they talked about?

 

From my first post "Calibration of the Grimmspeed intake was found to be 12-13.5% above the stock MAF curve (12% from 1-4 volts tapering up to 13.5% @ 4.1-5 volts). "

 

I think what wicknetzel is getting at is that you had indicated the K&N needed a ~2% MAF scaling in certain regions, and your most recent post indicated that for the GS CAI, you preserved the K&N scaling and added 12% on top of that (?):

 

"We installed the grimmspeed intake and I added 12% ontop of what was already added for the K&N panel."

[luchadorjose]

I would have loved to see the results of the car being maxed out in its tune with the stock airbox, compared to maxed out with the grimmspeed airbox. Otherwise it seems like you're comparing two artificially hamstrung parts

[wicknetzel]

Okay... just thinking more about what you did here and going to back to read your testing, I realized you never actually did a run with the GS box and posted the results as it was intended to be used.

 

If you look back at their design work, this is something they posted:

http://cdn2.bigcommerce.com/server5000/8fmm9av/product_images/uploaded_images/intake-subaru-0814-wrxsti-16-original.jpg?t=1430767128

 

So you aren't really testing the product as intended for use by sticking the stock ducting in and diverting air straight into one side of the filter. This is not allowing the air to flow well around the filter and could change results. Part of the design was also to isolate the MAF in the box.

 

I understand the temperature concerns, and do not disagree that it's worth further investigating options to fix that... but I don't think you can come out and say this:

 

Moral of the story is the GS intake could not beat the panel with its provided air box.

 

just based on the testing you've posted here... it's just not good science to make a claim without actually testing the parameters in question.

[solidxsnake]

Okay... just thinking more about what you did here and going to back to read your testing, I realized you never actually did a run with the GS box and posted the results as it was intended to be used.

 

If you look back at their design work, this is something they posted:

 

So you aren't really testing the product as intended for use by sticking the stock ducting in and diverting air straight into one side of the filter. This is not allowing the air to flow well around the filter and could change results. Part of the design was also to isolate the MAF in the box.

 

I understand the temperature concerns, and do not disagree that it's worth further investigating options to fix that... but I don't think you can come out and say this:

 

 

 

just based on the testing you've posted here... it's just not good science to make a claim without actually testing the parameters in question.

 

He did take runs with the airbox in place, but timing was being pulled due to high IAT. Because of that, it was impossible to dyno the results "as it was intended to be used," because it clearly wasn't intended to be used on a dyno.

[luchadorjose]

You guys had those air mover fans? The ones that look like snails? Maybe grimmspeed used those big propeller style fans, I know ECS near me uses one of those, it's about six feet tall. And EFI uses this homegrown solution, it's basically one of those air mover fans blown up to a box as big as a minivan, standing in front of it is like standing in a hurricane. There's so many variables in play

[solidxsnake]

You guys had those air mover fans? The ones that look like snails? Maybe grimmspeed used those big propeller style fans, I know ECS near me uses one of those, it's about six feet tall. And EFI uses this homegrown solution, it's basically one of those air mover fans blown up to a box as big as a minivan, standing in front of it is like standing in a hurricane. There's so many variables in play

 

We were at ECS Womp, I misread ECS as ESP =_=

 

The problem is that there's no good place to duct air into the fender well on the dyno without removing the bumper cover or the fender. And hell, if removing the box is enough to cause turbulence issues that makes the GS intake not operate as intended, I can't imagine what sorts of turbulence issue you'd get from putting a high-flow squirrel-cage fan pointed directly into the filter inside of a closed box

[Max Capacity]

I thought the test was done on ESP's dyno in Sterling MA, not ECS in Windsor locks ?

 

If it was at ECS I would have stopped by on my way home from work, about a mile away.

[covertrussian]

He did take runs with the airbox in place, but timing was being pulled due to high IAT. Because of that, it was impossible to dyno the results "as it was intended to be used," because it clearly wasn't intended to be used on a dyno.

 

The bigger question is why the timing was being pulled at only 100*F IAT's, that would be around 1* pulled since stock pulls 2* at 104*F, which is not enough to make a fuss about. If timing is being pulled due to knock then the base tune is too aggressive for this intake and you are way past MBT.

 

Now TA's graph does show that GS intake without box had more timing, which tells me that lack of box messed with MAF scales enough that it was running more to the left column instead of the same column as stock airbox. Which means the MAF scales needed to be upped a bit more without the GS box.

 

I wouldn't say it's impossible to dyno the results either, 100*F is really not that high, I hit 130-150*F with stock airbox while standing at stop lights this summer.

[doobaruGT]

He did take runs with the airbox in place, but timing was being pulled due to high IAT. Because of that, it was impossible to dyno the results "as it was intended to be used," because it clearly wasn't intended to be used on a dyno.

 

These can be adjusted. That's part of why so many are so skeptical.

 

1. Viable Data

One of the biggest points of contention seem to be that this was performed on a car that was previously tuned. There's not a great way to compare here since we're changing the tuner/tune in addition to the parts on the car. A proper method would isolate a single variable. I think TA attempted to do this with controlled parameters but still fell short of a good viable isolation since controlling for timing will skew results anyway; a new intake requires modification to the map as they know already. Isn't the point to optimize the map for each setup? Not to mention the ambiguity of parameters set for the existing tune. Why wasn't the car road tuned and logged instead or at least in addition to the dyno?

 

2. Airbox Design Objective

Recall that the primary objective behind the GS box was not only an attempt to isolate it from engine temps, but also streamline the entry around the cone and balance MAF readings. Having a proper MAF is worth its weight for those of us who know much in the way of tuning at all. While GS mentioned that was a goal, I only remember seeing some qualitative image showing the flow over the MAF sensor.

 

 

3. Real world...

Why not adjust tuning for timing if IAT's are causing timing issues? Again, this lends itself to the tuning objective- optimization. I would have liked to see TA run a baseline optimal tune for each setup and adjusting that optimal output per part addition/removal. GS provided some of that data here.

 

 

What I see here is a closed hood and a fan oriented towards the hood scoop. As solidxsnake mentioned, this was not intended to be used on a dyno, which, to me, seems to shoot this whole testing procedure in the foot. A dyno is for tuning and every effort must be made, as far as it depends on you, to model real world conditions.

 

In genuine driving conditions, we would have air flowing towards the scoop (I would hope, though I've always wanted to conduct some experimentation of this flow dynamic over the hood), into the bumper cutouts, and beneath the car, presumably pulling warmer air away from that stinking heat-soaking header. Barman mentioned the same thing earlier IIRC. Driving heat away/further isolating from the intake area is probably a great "step 2".

 

Intake Physics

TA mentioned that they gave considerable time between runs to prevent power loss from too many runs and to allow for a better baseline from which to draw conclusions. Good practice there, but I'm now thinking of the picture above. I'm imagining all that heat accumulating under the hood and then running a dyno. Maybe that was something that was adjusted for but wasn't shared? Did this vehicle have a splash guard? I can almost feel the header radiating up the fender well from here

 

In any event, IAT's are important to a degree. We all want to avoid knock. A better flowing intake is naturally going to have a slower flow velocity and contribute to (relatively) higher temps. I think Chase mentioned that a few times. I'm sure that delta in velocity is nominal.

 

Having said that, would you rather have a lower IAT with more challenging predictability in MAF, or a slightly higher IAT with an extremely predictable MAF? I only base this hypothesis on the hope that TA's IAT data is accurate. Speaking of which, TA mentioned about the delta in the driving IAT's and said they were similar but I might have missed the hard numbers.

 

Finally, it's important to remember that GS was going for a "Complete System" with MAF in mind (see wicknetzel's photo from GS above). It's trivial arguing about the IAT's alone, especially when measured this way. The scientific community always submits and yields to peers prior to publication for the sake of valid information. In the end, I think we're all on the same page here whether we realize it or not

[GrimmSpeed]

Hey guys - just getting caught up on this and there's some really interesting discussion. Chase is also now aware of the thread and will definitely have some insight to share tomorrow. Much of what I'm seeing here are things that we've seen and discussed internally while working through our design considerations and testing procedure. We're very well aware that scientific intake testing and the control of variables is a delicate thing, especially on a dyno. There are a few things here that are definitely worth further discussion, but Chase will be able to expand upon that.

 

I'm going to leave the technical response to Chase, since he's the expert on our intake systems, but I wanted to chime in with one request. I'm all for open and honest feedback, positive or negative, but please give Chase, the team at DB Performance and anybody else involved in our testing the benefit of the doubt. If you're going to question anything, question their knowledge, methods and design decisions, not their honesty and integrity. These guys were selected to do these jobs because they're true car guys and talented engineers, above all else. If there's one thing that I can absolutely promise you, it's that Chase would leave his career with GrimmSpeed before compromising his engineering ethics to pad results or mislead customers. Chase is passionate about designing killer car parts and his hair, not sales figures or the bottom line (much to my displeasure, at times).

 

If you know GrimmSpeed, you know that we don't sacrifice performance or quality for the purpose of getting to market sooner or any other nonsense. To be perfectly frank, the LGT community is an extremely small portion of our business and if we didn't feel that a product performed well, we simply wouldn't bring it to market. We're very good at a lot of things and have no interest wasting engineering, manufacturing, marketing, sales and customer service resources supporting products that we're not proud of.

 

To those of you with intakes yet to be installed, I'd strongly encourage you to do some additional research regarding proper dyno testing procedures in a scenario like this and continue following this discussion before scrambling to unload your kit. With that said, if you do decide that you're no longer interested in keeping it (as with any GrimmSpeed product, for any reason), we'll gladly take it back and sell it to somebody that will properly appreciate it.

 

Looking forward to keeping an eye on this and seeing what Chase and others have to add!

 

Matt Beenen

Engineering

[xtea]

I am so glad they will be adding their voice to this discussion. I feel like grimmspeed and chase can offer some help, thoughts and tips to get the most out of this device. I really want to be happy with it and my hope is that it really is better then "that other popular intake".

 

Thank You Matt for chiming in!

[fahr_side]

I wonder if the 08-14 WRX that this intake was originally designed for had more cold air flow to the fender.

 

I don't think there's more cold flow, but on my GH8 you cannot see the header from inside the fender area.

 

Back on topic I would like to see how the stock airbox compares to a cold air intake on a bigger turbo setup. It's always said that the stock box is good up to 300 wheel horsepower before it becomes a noticeable restriction.

 

 

In a few weeks I'll be installing a 1.5XTR on my GH8. I will be doing an STi manifold and intercooler swap to make that fit, so I'll be installing some 750s in gutted TGVs at the same time. To make it easier to dial in the injectors I'll use the stock intake at first, then add the GS intake to scale that. If I have time I will try to max out the stock intake on gas only and see if the GS one does better from there.

 

EDIT: My mechanic is too busy to install my manifold and turbo this week so I may just put the intake on and scale that first. I'll do a Virtual Dyno pull on stock intake first and see what differences there are on my current 18G setup.

[magnusonsubie]

As an engineering student myself I don't think I could ever see any engineer in the field sacrifice their reputation by forging data to help their design look better. I hope to hear from Chase about how their test compared to this test but I will still be holding onto my intake to install along with some other parts in the somewhat near future.

[Chris GTO TT]

After driving around today with the intake installed and watching Intake Air Temps I cannot agree with the claim the IAT's will be significantly higher than with the stock airbox/snorkle. As I have said in other posts here my observations show IAT's about 4-7 deg above ambient with this intake. The stock setup had IAT's of about 10-12 above ambient. Just like stock the IAT's will spike when you park or idle at a stop.

 

Ambient temp was 81 degrees per the Nav display on my spec.B, and the IAT was hovering around 85-87 degrees on the freeway at 70 mph. I then got off the freeway parked at a store for about 10 min. When I restarted the car the IAT showed 118, I pulled out into some stoplights and the IAT hovered around 96-98 degrees. Once I got into some contentious driving driving IAT's dropped to the mid 80's again.

[fahr_side]

As an engineering student myself I don't think I could ever see any engineer in the field sacrifice their reputation by forging data to help their design look better.

 

As an engineer with 30-odd years out in the field I can tell you with some sadness that you will get pressure from upstairs on a regular basis to fudge your numbers. This is why in places like Germany a product brochure is considered a legally-binding contract with the consumer and they can and will sue when things don't perform to spec. What you do is on you. I've left some quite lucrative jobs because I would not put my name on anything I wasn't confident in. Many more just roll over.

 

 

 

Sent from a device using some software.

[doobaruGT]

After driving around today with the intake installed and watching Intake Air Temps I cannot agree with the claim the IAT's will be significantly higher than with the stock airbox/snorkle. As I have said in other posts here my observations show IAT's about 4-7 deg above ambient with this intake. The stock setup had IAT's of about 10-12 above ambient. Just like stock the IAT's will spike when you park or idle at a stop.

 

Ambient temp was 81 degrees per the Nav display on my spec.B, and the IAT was hovering around 85-87 degrees on the freeway at 70 mph. I then got off the freeway parked at a store for about 10 min. When I restarted the car the IAT showed 118, I pulled out into some stoplights and the IAT hovered around 96-98 degrees. Once I got into some contentious driving driving IAT's dropped to the mid 80's again.

 

Thanks for this, Chris.

 

For other reference, I currently have an AEM intake and I took two different cruises, one last evening and one this morning.

 

I found IAT's at cruise/light throttle to be around +1-4*F. Yesterday the ambient temp was 60F according to my display and the AP displayed 63-64F consistently. The highest it reached was 75F after a red light, but it only took ~15s of moderate throttle (30-50%) for it to return to the delta 1-4F

 

This morning, I ran another cruise check with an ambient temp of 37F. IAT's were +1-3F with the peak at +19F after a red light. They returned to +1-3F within ~20s while applying 30-50% throttle.

 

I'm installing my GS intake this week so I'll be able to run some comparisons. It's not totally fair though since ambient conditions will be different by then, but at least it's still something.

[fahr_side]

... some contentious driving....

 

 

Haha. Nice turn of phrase there.

 

 

Sent from a device using some software.

[Chris GTO TT]

Haha. Nice turn of phrase there.

 

 

Sent from a device using some software.

 

Ha I thought it looked funny when I submitted it. I even spell checked it