Turning Vane in Intake Elbow vs. Without "Dyno" Results (2024 CX-5T)

[AL Cx5]

I made a turning vane in the tight radius elbow coming out of the air filter box on my 2024 CX5 Turbo. This elbow is smooth bore and discharges to a ribbed straight section. The next section is plastic pipe that transitions on the outboard side to a oval tube.

I had some scrap flooring and flashing material and made a rough turning vane. I went for a drive. My butt dyno said yes this works. I didn't say anything to my wife about this mod. She went about her daily driving around town. That night at dinner she casually mentioned the car had a lot more pick up. That confirmed my butt dyno "data".

I made a somewhat more refined TV and logged per my tuners instructions 2 WOT pulls using MazdaEdit. Using the Dyno function I compared this to a previous log. I go to the same spot to log WOT every time on a low traffic highway.

Here are the dyno logs. The HP and Torque numbers seem to be optimistic. This is like other Hp and Torque calculators I've used. It was interesting that peak torque was higher at a higher RPM with the TV and HP was higher at lower RPM without the TV.

I need to go back and log from a stop through 3rd gear with and without a TV to get a fair and accurate calculated dyno.

I like how much smoother the curves are WITH the TV. It appears our FE has improved as well.

Here is what Mazda published. Without TVs log dyno mirrors this more closely than with the TV. This confirms the TV impacted the performance.

 

[AL Cx5]

Here is what my first shot at a turning vane looks like in the elbow.

Using some scrap flooring and roof flashing.

It is staying in place, SO FAR. I carried around tools to remove it if it bloced the air flow.

I trust it to stay put to let my wife and grand kids go out at night. So far she has returned!

Here is the outlet. I need to trim the top piece a bit more.

Here is the inlet to the elbow.

 

[N7turbo]

Obviously there is a change but the numbers are wild. Which run came first? Same day?

There is apparently less power at "driving around town" RPMs.

There is something going on at 4,000 RPM on both runs. Power nosedives. I think the 4,500 RPM thing is an anomaly.

I'd be paranoid about my turbo sucking up a piece of that. I'm thinking finding someone who can 3D print would be the way to go.

 

[AL Cx5]

Obviously there is a change but the numbers are wild. Which run came first? Same day?

There is apparently less power at "driving around town" RPMs.

There is something going on at 4,000 RPM on both runs. Power nosedives. I think the 4,500 RPM thing is an anomaly.

I'd be paranoid about my turbo sucking up a piece of that. I'm thinking finding someone who can 3D print would be the way to go.

The logs are about a week apart BUT on the same stretch of road.

Other logs show the same anomaly. I have sent these logs to my tuner. If we discount the HP at 4500 RPM. TV is a clear winner. The tune closes the waste gate at 4k, thank you very much Mazda. MazdaEdit is missing some key tables for REAL tuning. Yet it has helped with power and FE...

Before I spend more time refining this I wanted to "see" some results vs. our butt dynos.

I'm encouraged the TV refined the curve and has more area under the curve. It runs much smoother as the curves suggest. The TV is putting more than half the air volume through the inside and avoiding two flow killers. The ribs in the straight section and the transition in the plastic TIP.

I'm sure you can agree, numbers don't lie. I hope this post will encourage another owner to build
a TV and refine the design. 3D print would be a winner. I might start looking into that as well. This is a prototype for sure!

This is not my first rodeo. We have been experimenting with TVs on our Cummins diesels for some time. We have MUCH more at stake than these toy 2.5ls. With that said, as I mod and tune this 2.5l, I'm more impressed with it. This week I completed solo and towing tests with some extra TVs in the TIP. I say between 10 to 15% gain in FE at partial throttle. I don't do WOT tests any more after breaking an expensive trans. And I'm not after MAX power but max FE on this diesel.

As side: my wife was asking last night what she would do if something happened to me with the CX 5. I said, well have the dealer reflash it back to stock and remove my bolt on mods. Her reply was I won't like the car any more!

 

[AL Cx5]

Diving into the data a bit more confirmed that a turning vane in the first elbow produced more power. Power also translates to better FE.

Looking at fueling and boost at WOT, same gear, nearly the same RPM shows the TV made more boost and added more fuel compared to without a TV.

I picked 2500 RPM as a place our cars operate often. AT WOT: Boost increased 16.9% and fuel increased 13.9% to support the demand.

Here is a screen shot of one of the tables I use to get the comparison.

 

[N7turbo]

I think the real test would be comparing to a straight pipe.

 

[AL Cx5]

I think the real test would be comparing to a straight pipe.

If you are referring to an aftermarket smooth bore elbow with extended straight section, it would be better than an oem piece.

Looking further down the tip, the part I modified discharges into a transition. Without a TV most of the air hits this transition which creates significant turbulence or pressure drop.

The gain to cost ration for this mod is the highest I have found. For the average owner, this offers a lot of bang for the buck. Of course one could redesign the elbow and or tip, which I have considered.

What I can always count on is N7turbo will find fault before doing the mod.

UPDATE: I found a 2.5 x 3.5 reducer 90 degree elbow and ordered this AM. Over on the other platform, we found that TVs help smooth bore fittings. I'll refine the TV in the smooth bore. I've not liked the ribbed straight section of this fitting from day one.

Looking further at the logs, at WOT the ecu didn't pull as much timing due to knock with the TV. Another indicator we are on to something.

 

[AL Cx5]

Two updates:

Looking at the MAF sensor with and withiut TV is a good indication if this mod is working. The MAF measures air flow through the engine. More air means a great chance to make more power.

I'm using logs from several weeks apart. No TV log was about 10 degrees cooler. At 4300 rpm the TV picked up 2%. At 4800 the TV picked up 7% or 18 HP.

2nd part:
I got the HPS 2.5 to 3.25 reducing elbow today. Vivid Racing had a good price and quick ship. I got the clamps off of Amazon. I trimmed both ends to fit and I stalled. Without the TV it was better than the OEM without a TV.

I moved the TV to the smooth bore 90 elbow and went for a spin. Low end torque jumped! Traction control light came on in 1st gear wot and it took off.

It makes sense a smooth bore is better. I was pleased with how much better.

I'll get some logs OEM vs smooth bore elbows all within an hour of each other to compare. This looks promising.

I would have preferred a black elbow. Vivid didn't have black in stock. I have less than an hour and $100 in this mod.

 

[AL Cx5]

Today I took my time and made a new turning vane for the smooth bore elbow. I made a WOT log and using MazdaEdit made a realistic dyno curve. The engine comes on so smooth and strong to 4,600 RPM.

The power comes on much smoother with a proper TV. I pulled out into traffic, the car accelerated quickly and smoothly. I'm very happy with this set up. I'll post pics of the TV and elbow.

Tomorrow, I'll take the OEM elbow and make two WOT logs, OEM and smooth bore with the TV.

Here is the curve. HP and torque that is close to DRTuned dyno results. DRT got 215 WHEEL HP, using 0.82 to convert to Flywheel HP is 262 HP.

Comparing air volume from the MAF senser and calculated % engine load is another valuable tool to compare one elbow to the other.

Here is shot of MAF and % engine load at 4669 RPM. If we use MAF value X 1.25 it comes up to 201 HP. This is more inline HP of this engine on a 95F day. Using MAF and engine load is a better comparison between the OEM elbow and the smooth bore/TV elbow.

Looking at DRTuned published dyno of a 2.5 turbo, it made 215 WHEEL HP or about 260 HP. The max MAF was 195 g/s or 244 HP. So things are not lining up perfect. I'm mostly interested in measuring the difference between the OEM elbow and the smooth bore/TV.

Here is DRTuned dyno. Note this car has several bolt on mods to improve HP.

 

[AL Cx5]

Here are some pics of the smooth bore with the turning vane installed.

 

[sinistriel@]

Today I took my time and made a new turning vane for the smooth bore elbow. I made a WOT log and using MazdaEdit made a realistic dyno curve. The engine comes on so smooth and strong to 4,600 RPM.

The power comes on much smoother with a proper TV. I pulled out into traffic, the car accelerated quickly and smoothly. I'm very happy with this set up. I'll post pics of the TV and elbow.

Tomorrow, I'll take the OEM elbow and make two WOT logs, OEM and smooth bore with the TV.

Here is the curve. HP and torque that is close to DRTuned dyno results. DRT got 215 WHEEL HP, using 0.82 to convert to Flywheel HP is 262 HP.

View attachment 329842

Comparing air volume from the MAF senser and calculated % engine load is another valuable tool to compare one elbow to the other.

Here is shot of MAF and % engine load at 4669 RPM. If we use MAF value X 1.25 it comes up to 201 HP. This is more inline HP of this engine on a 95F day. Using MAF and engine load is a better comparison between the OEM elbow and the smooth bore/TV elbow.

Looking at DRTuned published dyno of a 2.5 turbo, it made 215 WHEEL HP or about 260 HP. The max MAF was 195 g/s or 244 HP. So things are not lining up perfect. I'm mostly interested in measuring the difference between the OEM elbow and the smooth bore/TV.

View attachment 329845

Here is DRTuned dyno. Note this car has several bolt on mods to improve HP.

View attachment 329848

On newer vehicles the MAF conversion of 1.25 is a more difficult conversion and does not always work well. The dyno room was 90-95* during the 232hp run and the MAF flow at that point of the curve was 180g/s. Stock EVO X and Stis put down around the 230-240 mark on this dyno so that should give you a good representation of what that translates to.

 

[AL Cx5]

I'm new to Mazda/import tuning. I saw 1.25 factor and used it as a gauge if my logs made sense or not. it would be nice if 1.25 factor was spot on but that is asking a lot....

Tomorrow I'll log WOT with the OEM elbow and the smooth wall/TV el. I thinking the MAF will give a good comparison as a check to the MazdaEdit generated dyno graph.

Here is todays best pull, MAF peaked at 195 g/s. Using EVO X HP to MAF ratio, the 2.5l T made 251 flywheel HP at 4367 RPM. That would line up with DRT's modified 2.5T, SAE dyno number, 215 Wheel HP, 262 Flywheel SAE HP.

AND we haven't brought up timing or KC! I'm going these variables out for now until I have a better understanding how Skyactive controls timing and KC....

Thanks for a comparison MAF and HP. That helps!

 

[AL Cx5]

Today's test was sort of a bust. The TV twisted inside the elbow restricting flow. I removed it and made a WOT with the smooth bore elbow.

Smooth Bore Test:
At 4300 RPM in 3rd gear it made 196.82 g/s according to the MAF sensor. If we use the factor from the EVO, post 11, it made 253 HP or use 1.25 x MAF = 246 HP. The solid red line is MAF and dotted red % engine load. Above 4500 RPM engine load falls off. This is where the ECU opens the waste gate and reduces boost pressure.

It was a hot, 94F, 40% , 30. Hg pressure after noon. I'm about 550' above sea level for those who want to convert to SAE HP.

I might fashion a turning vane from sheet metal this week and try it again. I have enough data to compare.

I scoped out a lightly traveled 4 lane state highway to run partial throttle testing as well. It loads the engine enough to make it down shift along with down hill and flat sections. I'll use this to test partial throttle on this and other mods. Hopefully this provides some solid partial throttle data.

 

[skid00skid0]

air duct guide vanes at DuckDuckGo

DuckDuckGo. Privacy, Simplified.

duckduckgo.com

That long run of material will induce a lot of drag...

 

[AL Cx5]

air duct guide vanes at DuckDuckGo

DuckDuckGo. Privacy, Simplified.

duckduckgo.com

That long run of material will induce a lot of drag...

Good point. I need to learn how to suport the TV in the 90 in 2 places.

I'm looking at making a metal TV and supporting it with bolts cut to length. These will be nylon locking nuts on each side of the TV. I will likely double nut each side and use locktite. I don't want the turbo/engine to digest a nut or bolt....

All the mods are working on the CX5 after some start up issues. My wife has requested that I stop tinkering. This is 80% her ride.

I'm about to shift gears and weld in a TV in my 5.9 CTD intake elbow to the engine. This is on the pressure side of the turbo and might yield higher gains. If this works, I'll move back to a TV on the CX5.

I've ordered an air filter box for the CX5. I'm going to use what we learned on the CTD air box and add another air inlet to the box. It will have a back flow damper to meter the air in based on demand. I'll beg forgivness!

The smooth bore reducer without a TV looks like a winner. FE is up, about 2 mpg and less turbo lag. I need to get more seat time...

 

[AL Cx5]

Yesterday I tested rev 3 turning vane. This revision was made of 22 ga, steel with SS bolts. I wanted to securely mount the TV in the reducer. Rolling the leading edge around a thru bolt provides a smooth entry. I double nutted with locktite on the nuts and sealed with silicon.

Initial tests were underwhelming compared to an oem fitting. About a 1% increase in power at WOT. My road test on CC at 55 mph showed about the same results, 1%.

The road test is on a lightly used state highway. It has one steep grade the engine down shifts to pull. It takes a decent grade to make the turbo down shift. I noted where it down shifted on both tests. The aftermarket setup traveled further up the hill about 200 yards.

Last night we drove in to the city. I noted trip mileage on each leg. Inbound 27.5 mpg, return 25.7 mpg. The return is climbimg out of the valley. This is ruffled 50/50 city vs expressway.

Normally I get 23 to 24 mpg on this trip.

I might do an oem vs aftermarket TV mileage test over the same test course to confirm or not my drive last night.

 

[AL Cx5]

I had time today and after seeing the MPG on the short trip last night, wanted to see an A to B test.

I drove east bound on I20 from Birmingham, 65 MPH on CC. drove 17 miles to an exit and turned around. I stopped at the same spot on the entrance ramp, turned off the engine and zeroed out the B trip meter. Also stopped at the same spot and recorded the results.

Smooth wall with turning vane surprised me on the trip, going down a hill MPG went up to 48 MPG. The stock elbow/reducer only went up to 36 MPG. At that point I knew the smooth wall/TV was going to do better.

Results after driving the 38.2 miles

Smooth wall with Turning Vane: 36.0 MPG
OEM elbow/reducer: 31.6 MPH

Percent difference: 13.9%

The old adage, if it sound like it's too good to be true, then likely it's not. AND a hypothesis needs to be confirmed by another lab before it is accepted.

With the above in mind, I'm asking/encouraging someone else to do this mod. I'll guide you through the process. The only power tool needed is an abrasive angle grinder or other method to cut 22 ga metal and a few screws and a drill. PM me if you are interested and have a turbo 2.5l.

I can feel more low end torque with the smooth wall reducer with turning vane. We felt the same results on our Cummins Diesels with this mod as well.

 

[AL Cx5]

I'll leave this right here.

AutoSpeed - The Patent Files: Turning Vanes

 

[N7turbo]

They can certainly help with sudden directional changes, but they do next to nothing in a smooth, macaroni-style elbow (depends on radius/width).

Turning Vanes: Necessary Component, or Efficiency Reduction Device?

The installation of turning vanes in HVAC ductwork is perhaps one of the greatest sources of contention between mechanical contractors and HVAC engineers. Why? Because many mechanical contractors b…

buildingengineer.wordpress.com

 

[AL Cx5]

It's application dependent, for sure.

The close radius 90 exiting the air filter box, to a reducer on the outer radius is an application for a TV. The long radius elbow to the turbo is not an application.

Post #1 shows a nice gain in HP at WOT.

My FE has been very good after this mod. Last week my wife drove to the country that included stop and go traffic, state 2 lane highways, hilly rual and dirt roads. She got +33 mpg.

For us, WOT or more power is NOT our goal. FE is. The small gain in total powwe is icing on the cake.

I hope someone else with a turbo does this easy mod and post their impressions.