K & N air cleaner conversion with SU's

[stugilmour]

My 1971 Plus 2 is converted to SU's running in a Stromberg head. Presently the air cleaners are the small foam "domes" with a chromed metal mesh and a retaining clip. Don't have the car at home right now, but if required I can measure the SU throat size, but assuming they are 1,75's to match stock Strombergs.

Anyway, thinking of converting the the "sandwich" style K&N's to provide better air cleaning. Any advice on the correct model number/size to get? What have folks with Stromberg's installed?

Will I have to fabricate my own backing plate perhaps, or is this a pretty straight-forward bolt on? I think I know the answer here, but hope springs eternal!

I understand from Geoff Minors that one has to be careful with fouling the bonnet. Are the resulting air cleaner assemblies round or oval to prevent fouling?

I have looked over the K&N site but am lost in the choices.

Any help appreciated.

[steveww]

I use a Pipercross PX500 with the Strombergs on my Elan. They even do a back plate with the correct holes in it I also use SU type ram pipes to tidy up the flow. I can post some pictures if required.

[Lincoln62]

If you want to use pancake style filters you should be looking at ones that you soak with oil. Dry foam filters are not very effective. I worked at an engine shop many years ago and every engine with foam filters that we pulled apart had badly scored bores.

Peter
66S2

[steveww]

The Pipercross filter is an oiled foam type.

No scoring of the bore when I pulled the head last summer after 4 years with the Pipercross.

[bill308]

I recommend the factory OEM air box system with maybe a K&N conical air cleaner at the front end, replacing the OEM air filter. You get the benefits of a cooler charge and reasonable air filtration.

Bill

[steveww]

I did some tests comparing inlet air temp using a front mounted filter and direct to carb filters. At a steady 50mph there was nothing in it. Stationary at tickover the front filter was initially a little cooler but is soon caught up and then took longer to cool down again once on the move. There is enough fresh air moving around the engine bay at speed for a front mounted filter to make little difference.

If you are going with a front mounted filter, ditch the standard housing it is quite restrictive. Better to fit a cone type on to the end of the pipe.

Direct to carb filters give more noise. One other consideration as far as Strombergs are concerned, the standard air box is not so good and does not allow the fitting of ram pipes. Good ram pipes do make a difference, make sure they are full radius types.

[bill308]

Steveww,

You posted:

I did some tests comparing inlet air temp using a front mounted filter and direct to carb filters. At a steady 50mph there was nothing in it. Stationary at tickover the front filter was initially a little cooler but is soon caught up and then took longer to cool down again once on the move. There is enough fresh air moving around the engine bay at speed for a front mounted filter to make little difference.

Questions:

Interesting results Steve. Would you be kind enough to share the details of your test?

Specifically:

1. Instrumentation used?
2. Conditions of the day(s)?
3. Placement and type of sensors?
4. Approximate elapsed time of the drive/test?
5. Was the test route and driving style the same?

Both my Elan and 308 plumb ambient air to the carbs, presumably to supply cooler air to the carb intakes. I'm a little suprised that two manufacturers would have bothered to add the necessary hardware if there were no benefit. In the case of Lotus, Colin might have had the services of Mike Costin, a noted aerodynamicist type of his time and co-founder of Cosworth, to call on for advice. The question in my mind is did they get it wrong?

In the case of the Elan, many have been frustrated with the lack of local engine/radiator air venting. I believe Lotus produced a service bulletin directing one to cut holes into the front wheel wells to improve engine compartment ventilation. The carbs are situated high in the engine cavity, so should be subject the worst of the hot air. It seems reasonable that ducting of cool air from in front of the radiator should lower the temperature of the intake charge.

I have never performed a definitive test to prove or disprove the validity of this assumption, but may at some time. If I were to design a test of this claim, I'd employ a laptop computer driven instrumentation system employing a minimum of 5-thermocouples (tc's), with a sample rate of about 1-set of temperatures every second. I'd also preplan the test conditions and go to some pain to equalize the driving pattern. The test route would last in excess of 1/2 hour and I would try to make both runs as equal as possible, from a driving standpoint.

For individual air filters, I'd place the tc's as follows:

1. In the cowl intake
2. In the air between the inlets of cylinders 2 and 3, centered between their intake faces.
3. At the air filter/carb face interface, in the middle of the air stream, to cylinder 1
4. Same as 2, but for cylinder 3
5 Same as 2, but for cylinder 4

For the air box system, I'd repeat the locations of tc's 1-4, but airbox/carb interface would replace filter/carb interface and I'd add an intermediate tc 5 at the entrance to the air box, in the center of the flow stream.

Once the data was acquired, I'd be looking at the air temperature rise, above the control tc1, in the front of the cowl. All temperature differentials would be referenced to this cowl tc. This temperature rise approach mitigates changes in ambient air temperatures and ensures comparability of the data.

Expected results:

For the case of the individual filters, let's call it case A, I'd expect the measured temperature differentials (tn-t1) at tc's 2,3, and 4 to be significantly larger than tc 1 and be comparable to tc4, in the engine bay.

For the case of the air box system, lets call it case B, I'd expect the temperatures differentials of tc's 2,3,and 4, to be noticeably larger than tc1, but less than in case A, and less that tc4.

Thermal testing is a tricky business, but performed with care, can yield useful results. If the temperature rise results are within a few degrees, then I agree, the ducting system will yield virtually no gain in perfomance.

Bill

[steveww]

I used 2 probes (see www.maplin.co.uk part number FE33L) one just behind the front grill to pick up ambient temperature the other in between the two carbs inside the airbox/filter.

The test was a steady drive at 50 mph down a quiet road near my house. The two runs were done less than an hour apart on the same day. I noted the difference between ambient and intake temp. Again there was nothing in it between the two arrangements.

One major difference is that my S4 has the radiator positioned right down in the nose, see http://www.steveww.org/index.php?id=26 may be this arrangement allows more air passed to the carbs.

If you want to conduct a more scientific study please do and publish the results here for us all to enjoy.

[bill308]

steveww,

Thanks for the details of your test. I hope I didn't come across as being too anal in my post but I do this sort of thing as a living, analyzing and validating the temperature effects on aircraft engine mounted electronics.

Your test was a good first order attempt at shedding some light on this issue. It may be that under normal driving, there is little or no noticeable benefit with a cold air box.

What was the temperature differential between the ambient air sensor, in the nose of the car, and the locations of interest?

Discussion

I looked at the device you used and came to the conclusion that it was not quit right for the task. I can't tell from the online information if it is thermocouple (tc) or thermister based, but I believe it is thermister based. From an absolute accuracy standpoint, thermister based systems are generally more accurate, as their signal output is much larger than that of a tc. This fact means that the electronics can be simplier, as the necessary amplification can be much more modest. TC's, on the other hand, give raw signals in the low milivolt range, at least an order of magnitude or two, less than a thermister and require cold junction compensation at the other end of the wire. Fom an electonics standpoint, more challanging and more expensive.

The size of your probe is quite large and massive for this particular measurement. It will be good for average readings as it will be somewhat imune to short duration chamges in air temperature. This is because the surface area to mass ratio is relatively small for your device so it will be slow to respond to dynamic changes in the air temperature. Your device would likely be very good for liquid temperature measurements though.

When we do a test that must measure air temperatures, we generally use 30 awg thermocouple wire, solder the sensor leads together with a single cross, and strip and spread the wires into a more or less diamond shape. The resulting wire/bead mass is very small and the surface area of the stripped wire, relatively large. This ensures the temperature of the thermocouple will change relatively quickly, to follow the air temperature. The down side of thermocouples is that they are not as accurate as a good thermister from an absolute temperature standpoint. Wire alloy variations, even using special limits wire, vary about +/- 2C from absolute true, for a given spool of wire. However the relative temperature readings between tc's is quite good if taken from the same spool of wire, provided the amplyfiying electronics are also good. An instrument grade amplifyer must be used. The bottom line is that one needs a fast response system for transient air temperature measurements.

The other shortcomming of your test is the test condition itself. You ran a near steady state test under which one might expect little difference between the sepatate air filters and cold air box. At a steady 50 mph, the engine is turning somewhere around 2800 rpm and the butterfly openings are pretty small. This means that the air flow into the cabs is relatively small, as is replacement of the air. In the case of the cold air box system, air is not being purged very rapidly, so air entering the system moves relatively slowly and has the opportunity to pick up some heat from the surrounding engine bay, by a convection/conduction/convection process, through the plumbing. If the test were run with with a heavy foot, air in the cold box assembly would be purged more quickly and the engine bay might be hotter, so perhaps one would see more of a difference.

You are probably correct in that your radiator forward mounting scheme may aid in keeping the engine bay at a somewhat lower temperature, possibly better ventaliation if you will, so the differentials between the two configuration were small.

Thanks for sharing you test results with us. I'll put the more comprehensive test you suggested on my to do list and I will share the results.

By the way, thanks for the link to your web site. Lots of good info there, particularly the section on porting. What was the material you used to cast the port profiles?

Bill