Brickley Engine: BSFC comparisons

[boonebucker]

Below are a couple of links to BSFC numbers for engines with similar technology criteria and displacement per cylinder. Does anyone have any numbers that would indicate for this type and size engine that these numbers are incorrect? The first link shows a sweet spot of about 440g/kW hr.for 179cc engine. The Brickley Engine is 173cc per cylinder and I am obtaining 350g/kW hr. The other link indicates numbers much higher than 440 g/kW hr and they will have to be converted from lb/Hp hr. I used the 3/4 throttle numbers. Your thoughts?

https://online.tugraz.at/tug_online/voe_main2.getVollText?pDocumentNr=488002&pCurrPk=73453

https://www.arb.ca.gov/msprog/offroad/sore/swri-final-report.pdf

 

[BrianPetersen]

Engines that are calibrated rich (2nd paper) will show a BSFC that is highly dependent on how rich. These little engines essentially use fuel as part of the cooling. The cooling fan (air-cooled) will be a significant parasitic loss. These small engines typically don't have particularly optimised intake, exhaust, and combustion systems. Usually the compression ratio isn't very high (in the interest of the engine protecting itself from detonation) and the ignition timing is fixed - which can only be a compromise between not advanced too far for starting, and not advanced far enough for good performance under load.

Better-optimised gasoline engines usually have a best-efficiency operating point in the 250 g/kWh range, but those engines would have bigger cylinders, better-optimised fuel delivery and combustion systems and ignition timing, etc.

 

[boonebucker]

Thanks, Brian. I guess what I am wondering is if anyone might have any numbers from tests they have conducted themselves or have references for otherwise to demonstrate that the numbers in the links above are incorrect somehow. Granted all of these engines are running rich with $9.99 carburetors, mine included. In fact my engines may be the least expensive of the bunch.

 

[gruntguru]

Honda GXV160, fairly old technology. 163cc, 250g/hp.hr, (333g/kW.hr). Link

When comparing BSFCs, the big differences will be in combustion and thermodynamic efficiency. You need to focus on comparing FMEP if that is where your advantage lies. You should not be trying to match IMEP with the big players. Putting their tech on top of your low friction mechanism is the easy part.

je suis charlie

 

[RodRico]

When comparing BSFCs, the big differences will be in combustion and thermodynamic efficiency. You need to focus on comparing FMEP if that is where your advantage lies. You should not be trying to match IMEP with the big players. Putting their tech on top of your low friction mechanism is the easy part.

Agree 100%. Comparing FMEP isn't easy, however. I have long experience selling new tech to customers, and it's always a challenge to convince them the new tech is fairly compared to the tech it replaces. In my opinion, the trick to selling this new engine configuration will be an apples-to-apples FMEP comparison (same cylinders, pistons, rings, bore, stroke, friction coatings, etc) between the new Watts Link configuration versus a traditional crankshaft configuration. Fortunately, the new engine is derived from an existing engine. Unfortunately, I don't think the existing engine is four cylinder. Though it may be impractical at this point, it seems like the ideal configuration for apples-to-apples comparison would consist of a new four-cylinder boxer engine crankcase housing a two throw crank (flat-four) with the same top end assemblies mounted on the crankcase. This would allow direct apples-to-apples comparison of FMEP, torque, power, and BSFC between the standard configuration and the new.

 

[boonebucker]

Agreement here as well. I had seen the 250g/Hphr but have made a point to shy away from single numbers and prefer a family of numbers as a general rule. There are so many allowances regarding J1349, J1995 etc. and testing without air filters and mufflers I don't really know what I am comparing. Of course this is exactly your point. And as well why I suggested the links I did if comps are relevant to whatever degree. Something I found interesting about the manual link is that the BSFC for both the 118cc and the 163cc are the same. Also, I'm wondering about the flatness of the torque curve. It appears they used the maximum torque number for both engines to get the Hp number @3600rpm.
In any event, I am complete agreement with the focus needing to be placed on FMEP.

 

[boonebucker]

Along the lines of FMEP, once again it's about what a family of numbers points towards. Check out Figure 13 of this SAE paper.

https://pdfs.semanticscholar.org/f9c7/4733a470b57d42dc9d3180ef8ebad83d2656.pdf

While there is only one point close to my engine (692cc), the number for my engine @1500 rpm is off the chart literally, half-way between 0 and 5.

 

[RodRico]

I’m not too sure under what conditions the values in figure 13 were taken, are you? They could be calculated from dyno data under load. The only thing I can say with confidence is that Honda took comparable measurements on the new engine in order to compare it with their new engine and found a 10% improvement.

 

[boonebucker]

Good point. I might be wrong, but I took Section 7.4 to include Figures 12 and 13. Section 7.4 talks specifically about engine motoring with 0-20W. While Figure 12 refers to only two engines, Figure 13 refers to over 30 engines. I would think that the data for motoring complete engines might be something easily accessed for comparison for a manufacturer. I'm not sure what they might be comparing otherwise in Figure 13.

 

[RodRico]

I'm not sure what they might be comparing otherwise in Figure 13.

Figure 13 may show FMEP results calculated on a running engine using dyno results augmented by in-cylinder pressure measurements.

From HPWizard.com page on FMEP: "Experimentally, the power taken by the friction between the engine's components ( Pf ) is found by substracting the measured power at the crankshaft (P) from the potential power that we can obtain from the measured gas pressure acting inside the cylinder(s)"

From "Methods to Measure, Predict and Relate Engine Friction, Wear and Fuel Economy": "Motored dyno tests allow for direct measurement for FMEP. However, motored dyno tests don’t duplicate a fired engine conditions due the absence of ring loading from cylinder pressure and thermal effects... Fired engine dyno tests best model real world operation by introducing the cylinder pressure and thermal effects missing in motored friction tests. Fuel consumption will be measured. FMEP will be inferred from cylinder pressure measurements."

Installation of in-cylinder pressure sensors is routine in professional engine development efforts, and it's the only way to get a true FMEP measurement. Based on the number of reports I've read, motoring friction is apparently pretty effective in assessing lubrication, however.

 

[boonebucker]

I'm making some assumptions, however, would it not usually be the case that if one was dealing with FMEP, and not friction torque, the choice in measurement units would not be Nm. I'm not sure if I have ever seen FMEP converted from kPa back to a torque in Nm for a paper, but once again I could be wrong. Interestingly, it's not that uncommon to see Nm converted to kPa.

 

[BrianPetersen]

The normal case in a research environment would be to express it as a MEP ... but that "paper" was more like an advertisement, so consider the audience.

 

[RodRico]

Would it not usually be the case that if one was dealing with FMEP, and not friction torque, the choice in measurement units would not be Nm.

Maybe.

Fired engine friction is calculated from torque and RPM (measured on a dyno) augmented by measured cylinder pressure. These three variables can be expressed in many ways. It wouldn't surprise me at all if a given engine shop or research team elected to express friction in terms of torque, as cylinder pressure that wasn't expressed as torque, or in terms of Mean Effective Pressure[1]. You're facing the problem with using data collected using methods that aren't described. I can't tell you how many times I read a paper to mean one thing because it resembled what I expected (or hoped for) only to later discover there was no actual correlation between their result and my expectation. Master and Doctoral theses describe methods and results in detail, and they've helped me resolve faulty interpretations many times. I'm sure there are some that unambiguously describe friction meaurement methods and measured results. Perhaps if you comb through the results from a search like "engine friction site:*.edu filetypedf" you'll find some helpful theses.

[1] In my opinion, the best measure for comparison across different engines is Mean Effective Pressure (IMEP, PMEP, FMEP, and BMEP).

 

[boonebucker]

I'm thoroughly enjoying the thoroughness. Speaking of Master and Doctoral theses, I don't remember any comments on the one that I had posted in the previous thread.
I'll post it again:

https://scholarsmine.mst.edu/cgi/viewcontent.cgi?article=8100&context=masters_theses

 

[GregLocock]

Typically back in the day FMEP was inferred from the flywheel torque, rather than measured directly.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[gruntguru]

FMEP=BMEP-IMEP but the great difficulty is getting an accurate measure of IMEP. Minute errors in the location of TDC on the pressure trace produce large errors in IMEP.

je suis charlie

 

[GregLocock]

I think they were usually motored tests, for FMEP. I don't have Heywood to hand, what's that line of flywheel torque vs (?) fuel consumption called? that seems to me to be robust approach. Willans line, figure 13-5 in my copy I just found.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[boonebucker]

I posted the wrong link in my very first post. It was missing the most important part; Appendices A, B, etc.
Here is the link I meant to send:

https://ww3.arb.ca.gov/regact/sore03/app_g.pdf

I am compiling a spreadsheet with similar units converting lbs/Hp hr to g/kW hr.

 

[boonebucker]

The attachment below compiles the data from the two links I mentioned at the start of the thread.

I converted the numbers and plotted them as fuel rates (g/hr) on the ordinate and power (watts) on the abscissa. A horizontal line can be drawn at any fuel rate to compare the power developed by each engine. The two rotational speeds chosen were 2600 rpm and 3060 rpm. The availability of the data was the basis of the choice.

The displacement and the fuel rate (FR) number is used to identify the engine and run. The FR number is highlighted in yellow at the bottom right corner of the run.

For the 179cc engine, the BSFC numbers were worked backwards to the fuel rates.

Regarding comparison with my engine, the individual cylinder numbers (for 173cc) were determined by dividing the actual numbers by four. Interestingly, the assumptions made in doing this division direct attention to the exact areas where emphasis is being made regarding reduction in friction.

There is only one point for the Brickley Engine on the 3060 rpm plot because that’s all that has been completed to date. A plot similar to the @2600 rpm plot is expected.

Noteworthy is that in comparing the best engine tested (163cc) with the BE @ 3060 rpm the difference in output is 568 watts (1843-1275). This difference @2600 rpm, the closest point, is about 425 watts.

 

[boonebucker]

I thought I downloaded 3 attachments. Where do they show up if I want to see them?