I recently acquired a gutted airbox with K&N filter, and wanted to compare it with the stock intake setup.

Test vehicle:
98 LeSabre in sig (n/a L36)

Test conditions:
1) Stock intake (airbox and accordion) with new AC Delco 1096C air filter
2) Gutted airbox (inserts removed from both halves, front panel removed) with freshly cleaned/oiled K&N filter, IAT relocated into rear half of box, accordian replaced with smooth 3.5" rubber pipe coupler

Test method:
Calculate and compare WOT volumetric efficiency (VE%) in 4000rpm to 4500rpm range using capture data from AutoXray Scantool, 4 replicates per test condition (2 reps per day for each condition, done over 2 days).
Car warmed up at least 5 minutes before each run, to ensure coolant up to temperature (180-182 F) and car into closed loop before punching it. Outdoor temps for all runs 68-70 F.

Results:
Test condition 1 (Stock):
Run VE%
1. 71.4%
2. 67.1%
3. 69.7%
4. 70.3%
Avg 69.6%
SD 1.8%
%RSD 2.6%

Test condition 2 (Gutted/K&N/coupler):
Run VE%
1. 67.1%
2. 71.0%
3. 67.8%
4. 69.7%
Avg 68.9%
SD 1.8%
%RSD 2.6%

The VE% data were remarkably reproducible for both test conditions.

Conclusions:
On my motor, the gutted airbox/K&N filter was not significantly different than the stock airbox at rpm'* near (and just above) the L36 advertised torque peak. The gutted box clearly is capable of more flow than the unmodified stock box, so this indicates that any limiting restriction to air intake is elsewhere (MAF screen, TB bore, UIM runners, LIM ports, head ports, valves).

Considering the awful noise the gutted box makes on cold startup and its lack of measurable improvement over bone stock, I see no reason to use it on this engine unless further mods somehow make the stock intake the limiting restriction.

I believe that the lower LIM ports are currently the worst intake restriction on this engine.

Thats pretty interesting.
I wonder how much the VE changes in the upper rpm ranges with the gutted box and k&n.

I also am curious about that.

Good info though!

It is impossible to say. On every test run, regardless of test condition, the MAF reading hit a plateau above about 5200 rpm (i.e., MAF reading didn't change from 5200 to 5700 rpm). Thus, 5200 rpm seems to be the limit of "breathability" in my (nearly) stock L36.

Having had a look recently (on my Mother-in-Law'* 98 Olds 88 ) at L36 head ports, LIM ports, UIM runners, and TB bore, (and looking up the intake valve specifications), I have to suspect that the LIM ports are the limitation. I measured the LIM port area at about 1.3-1.4 sqin, compared with about 1.7 sqin for a head port (it was hard to measure the LIM ports, because they were so irregular in shape).

I bought a LIM from Ed Morad off EBAY, and I will repeat this experiment after I port that LIM and install it (probably 3 to 4 weeks from now). What I hope to see are:
1) higher VE% figures at 4000 rpm and above
2) MAF reading "plateau" occur at higher rpm (or disappear--i.e., occur at rpm>5700, beyond limits of my current shift point)

(I also bought a UIM off Ed, because I have some ideas about modifying the runners, too)

I hope I do NOT experience significant reduction in low-rpm VE% and low-end torque with the ported LIM. If I do, I'll probably go back to the unmodified LIM while I figure out my next step.

You'll find the head ports to be 0.9" by 1.9" on the L36. The LIM ports are so ugly, you can't even begin to measure them.

If you want pictures of a ported LIM for reference, I can post some tonight. I have it at home (L36). The first DIY L36 LIM porting job on this Forum had great 'seat of the pants dyno' results, but no scantool characterization. The second one (on my workbench at home) will be fully tested for flow on the car (before/after).

It'* important to remember the sequence of operation of the engine when thinking about this. You can increase the volume of air entering the cylinder by port-matching the LIM to the heads, but you also have to remember that only one intake valve at a time is fully open.

You need to calculate cam lobe lift and duration for all 6 cylinders to determine how much can flow at any given time, then compare that total cross-section to the smallest diameter upstream and make sure that'* enough to flow for the demand. Velocity has a big role in all this.

Try a search on this with regards to the L36 LIM ports. It'* been mentioned more than once. That'* why we're already working on it behind the scenes.

willwren,

I would appreciate some pics of the ported L36 LIM...especially if you have both before and after.

I know I don't want to kill air velocity by over-porting. The base "rule-of-thumb" as I understand it is: intake port cross-sectional area (CA) wants to be no more than 80-85% of the valve CA, so as not to destroy velocity. (If I'm way off on this, please let me know)

Rough numbers here:
Current LIM has port size roughly 1.35 sqin.
The LIM ported to 0.9" by 1.9" would be CA=1.71 sqin.
The intake valve (at 1.8" diameter, with estimated 0.3" stem diameter) has CA=2.47 sqin.

CA(LIM) / CA(valve) * 100% = 1.35 / 2.47 *100% = 55%
CA(ported LIM) / CA(valve) * 100% = 1.71 / 2.47 *100% = 69%

Even ported, we're well under 80%; however, the porting increases the port to valve ratio significantly (by over 25%), so there should certainly be a difference.
My expectation is greater flow at all rpm, but lower velocities at low end (thus some possible low-end loss), with max velocities at higher rpm than currently (thus high-end gain).

I'm running the stock PCM, cam and lifters, and don't plan to change any of that anytime soon (well, maybe lifters and PCM in the future ).
Will my FSM have the stock cam specs? (Or where can I find them?)

The cross-sectional area of the port decreases to ramp up velocity. I found that if you were to gasket match the LIM outlets, the port'* cross-sectional area, only in the LIM, decreases roughly 29%.
(topic: Curious about anything L36 UM or LIM?) I can't link it at the moment.
So, with an unported head and LIM to match, there would be even greater velocity.

Probably time for Mike1995 to post his 'test drive' results.

jr's3800 should be testing another later this month or sometime in March.