Piston IC Engines, Internal Inertial Loads

[FredRosse]

"The greatest load on a connecting rod is inertial load……………

I have herd this statement several times previously (that inertial loads are the big item, 5-10x the pressure loads), but as far as I can see with real calculations, this may be an old wives tale, or only applicable to screamers like chain saw engines and race engines?

From various forum threads, including eng-tips.com:

"…..keep in mind that only about 15-20% of the load in an engine is due to the compression load and the load imposed from the fuel firing on ignition. The majority of the load on the con rod is due to the forces developed by acceleration of the piston mass moving up and down the cylinder. "

"The greatest load on a connecting rod is inertial load, occurs at top dead center on the exhaust stroke, and produces a tensile stress, not a compressive stress. This is true regardless of engine load, and depends ONLY on the RPM for a given engine."

The eng-tips forum thread, in supporting the high inertia loads statement listed a 350 Cvevy engine at 5000 RPM, with 1.6 pound pistons, 3.48 inch stroke, showing inertial loads far above BMEP x Piston area. Further examination shows a result of the inertia loads calculated above were high by a factor of 32.2…..hummm.

I then went into the basement, where a small Diesel engine piston and rod could be found and weighed. Calculations based on the engine's rating again show the Inertial loads as a small fraction of the piston pressure loads

Since my calculations for real, and once very popular engines show something quite opposite what was stated, I request any real data to shed light on this issue. Not anecdotal evidence, but the real story from an engineer conversant on this issue.


Thanks in advance

 

[ivymike]

it depends on the engine and the operating speed. A turbocharged diesel engine is going to have much higher peak cylinder pressure than a naturally aspirated engine (gas or diesel), and typically a much slower maximum rpm. A slower 2-stroke diesel engine may not ever have a tensile load in the connecting rod - the load can be compressive at all times.

A couple of things to keep in mind:
1) max pressure load is not BMEP x area, it is peak cylinder pressure x area
2) rated rpm is not the highest rpm to consider if you have a transmission and wheels (or in other cases where forced overspeed is possible) - it is not uncommon to design a diesel to allow the engine to be forced to 50% higher rpm than rated (so a 2100rpm rated engine would have rods designed for >3150 rpm and might see that speed at some point).
3) 32.2 is a "special" error to have when calculating acceleration w/imperial units ... you might want to check that your unit conversions were correct.

 

[SomptingGuy]

A plot of (full load) 2E flywheel speed variation for an I4 gasoline will generally be a "V" shape, with the minimum at c. 2000-3000 rev/min. Below this speed, firing torque is dominant. Above it, inertia torque is dominant. This is even more clear if you plot each cycle's speed variation through the cycle in a 3D plot vs mean engine speed for that cycle.

- Steve

 

[FredRosse]

Yes, it is understood, all of the points you have correctly made here, inertial loads generally follow a second power function of RPM, cylinder pressure loads not, also:

1. BMEP is just the pressure as if it were acting throughout the power stroke without friction, whereas actual peak pressure is higher because of friction, and several times higher due to the gas laws, approx. PV^gamma = CONSTANT, so if it is shown that BMEP piston load exceeds the inertial load, then real peak pressures loads are far higher than the inertial loads.

2. Yes, engine RPM often runs beyond the rating points.

3. We must always be sure to respect dimensional analysis, and that all equations in all of engineering and science must be dimensionally consistent. I mentioned the "32.2" because this is a common mistake. Rusty calculators often forget that F = M * A in the English system requires Slugs as the mass unit.

If we are using POUNDmass units, then the following non-dimensional unity is of interest when applying Newton's law:

1 (ND) = ( LBf - sec2 / ( 32.174 ft - LBm )

Also when dealing with rotational dynamics, radians per unit time is generally appropriate.

 

[FredRosse]

SomptingGuy, OK for torque, but what about connecting rod loads, at the wrist pin and at the crank end?

 

[ivymike]

BMEP is just the pressure as if it were acting throughout the power stroke without friction,

Not at all... BMEP is Brake Mean Effective Pressure, and it is work per cycle divided by displacement per cycle. It is an indicator of how heavily loaded an engine is. It includes friction. It cannot be used to calculate peak cylinder pressure. With a complete cylinder pressure history along with the mechanical geometry of an engine you can calculate BMEP.

BMEP is a mathematical construct and can only be calculated, whereas cylinder pressure is physically real and can be measured directly.

For example, an engine I helped design has a cylinder pressure limit of 186 bar. We try to come close to that limit with our highest intermittent ratings. Due to exhaust temperature limits we were in the past limited to a rating which had BMEP=25 bar, but with a fancier turbo and wastegate we were able to keep exhaust temperatures in limits and achieve 28 bar BMEP. We actually ended up with about 4% lower peak cylinder pressure in the second case than in the first once emissions limits were met. So to recap, pcp was about 180 bar and BMEP was either 25 or 28 bar depending in which rating you calculated it for.

 

[FredRosse]

from ivymike: "Not at all... BMEP is Brake Mean Effective Pressure,.................BMEP is a mathematical construct and can only be calculated, whereas cylinder pressure is physically real and can be measured directly."

I agree completely, and I was only pointing out that BMEP is much less than actual peak pressures. I never intended to imply that peak pressures could be calculated by knowing the BMEP.

The thread is looking for data about connecting rod loads, inertial vs. pressure force.

 

[ivymike]

ok, so for the engine I mention above, the following apply:
rated speed 900 rpm
max overspeed 1250 rpm
cylinder bore 280 mm
stroke 300 mm
rod length 600 mm
approx. recip mass 60 kg
max cylinder pressure 186 bar
TDC force due to acceleration at overspeed: 18100 N
TDC force due to acceleration at rating: 9400 N
Force on piston due to gas @ PCP: 1,145,300 N

So max gas force >> max inertia force

 

[ivymike]

at rated power, the exhaust backpressure is in the neighborhood of 5 bar, which means that at TDC of the exhaust stroke (when gas force is approximately minimum) there will be about 31,000 N gas force (compared to 9,400 N in the opposite direction due to acceleration). You have to come off of peak power by a fair bit, so that boost and backpressure both drop, before you start to see any tension in the rod at TDC. You do, of course, see tension at TDC-exhaust when the engine is at rated speed with no load.

 

[Tmoose]

Attached are a few images from Chrysler's 1967 SAE paper about the "new hemi head high performance engine."

Fig 20 is the rod journal (not the I-beam) loading at 7200 rpm. The lightest Hemi piston weighed a whopping 813 grams, and as a result the conrod was pretty beefy too. It took several iterations of the big end and bolting arrangement to prevent big end deflection and overcoming the bolt clamping forces by "the inertia load of the piston and rod itself" at exhaust TDC at high rpm. 16,000 lbs in Fig 20.

Fig 9 is the #4 main journal loading at 7200 rpm. The paper says the middle 3 mains were "cross bolted" specifically to better resist the horizontal loads.

 

[Lou Scannon]

As alluded by Tmoose's post and attachment, the critical case for inertial loading in tension occurs under low pressure, max rpm conditions. Low [cylinder] pressure is experienced by all engines at TDC during overrun (motoring), and by normally aspirated engines on every other revolution, i.e.,during TDC exhaust/overlap. As mentioned above, it depends on the individual engine/application, whether the maximum inertial or maximum pressure loading is more severe from a stress point of view.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[FredRosse]

ivymike, Thanks for some real data.

tmoose, I could only get one figure, but it appears the hemi data at 7200 RPM is more than 46 years ago, and is a race engine, yes? I thought most normally produced relatively big displacement V-8 engines (USA Automotive) run around 4000-5000RPM?

 

[Tmoose]

I'll try Fig 20 again.

Yes the data is for one of the highly developed "race" hemi engines.

I believe you'll find that Cracks that develop in connecting rods in lesser service are the result of tension loads.
Here is an example of a lowly passenger car inline six that was the victim of details overlooked by either the design or manufacturing group that made it susceptible to death by fatigue by repeated tension loads.
Pages 449 and 450 and Figure 14 here -

http://www.studebaker-info.org/SAE/TSVE/tsvep449.jpg

http://www.studebaker-info.org/SAE/TSVE/tsvep450.jpg

http://www.studebaker-info.org/SAE/TSVE/tsvep453.jpg

Conversely cracks in crankshafts generally initiate in the fillet of the rod and main journals in regions that suffer most under compression loads, combined with torsional loads.

http://www.bmridersclub.com/images/Broken 5 Crank.jpg

http://www.forddoctorsdts.com/uploads/gallery/album_5/med_gallery_1_5_64705.jpg

http://www.evolution-fea.com/pics/s...ses_I4-crankshaft-dynamics-stresses_Iveco.jpg

 

[HarleyE30]

In a vast majority of engines, gas forces from peak cylinder pressure predominate stresses in the connecting rod. However, connecting rods tend to fail on tension, as it is far easier for cracks to form and propagate in tension. Rods that have failed in this way are uncommon now due to simulation methods. Even before computer simulation, simple stress calculations based on the cross sectional geometry of the rod should have provided enough information to design rods which won't break in normal service. At engine speeds exceeding the engine's capabilities, then the valvetrain would most likely give up first (usually valve float leading to valve/piston contact).

For rod analysis, the following operating conditions are typically investigated:

* Max torque (with inertia relief) (compressive)
* Max RPM (with gas force relief) (Tensile)
* Max overspeed RPM (with or without inertia relief - depending on who's doing the analysis) (Tensile)

Also, assembly loads are often added. Analysis at the big end (ovalisation is important for bearing analysis) is critical as is analysis for the connecting rod bolts.

 

[tbuelna]

With connecting rod fatigue, it's the net result of the inertia and combustion loads and what degree of load reversal (if any) is produced on the conrod beam throughout a full engine cycle, plus any additional loads such as dynamic bending or torsion, over a typical engine lifecycle.

An extreme example is that of a highly boosted, low speed 2-stroke diesel. The conrods in these engines are always loaded in compression. In fact, the pistons of these engines are often made heavier than they need to be, using piston inertia to offset the peak combustion pressure loading on the rod bearings.

 

[Lou Scannon]

In fact, the pistons of these engines are often made heavier than they need to be, using piston inertia to offset the peak combustion pressure loading on the rod bearings.

What a windfall for the piston designers! How often do you get to design something for high mechanical stress where the mass constraint is soft or non-existent?

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[FredRosse]

"How often do you get to design something for high mechanical stress where the mass constraint is soft or non-existent? "

Many places, commonly almost all static structures (not counting seismic building design) such as simple support structures. For articulated machinery, also plenty of applications where inertial loads need not even be considered, such as backhoes and bulldozer mechanisms. My sidewheeler beam engine turns only 60 RPM, no inertial dynamics to consider there.

 

[Lou Scannon]

Good examples Fred, thanks... as we are in the Engine and fuel forum, I neglected to express my thought process adequately... "by something", I was thinking chiefly of engine parts, i.e. those in motion and/or bearing the reaction loads from the cranktrain and valvetrain parts.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[turbomotor]

Interesting comment from tbuelna. The Wartsila RTA96 is a large two stroke engine with rated speed of 102 rpm and a stroke of 2.5 meters. Mean piston speed is 8.5 meters/sec. It would be interesting to know the relative impact of the inertial loads and the gas loads on the 300 ton crankshaft.

http://en.wikipedia.org/wiki/Wärtsilä-Sulzer_RTA96-C

 

[tbuelna]

turbomotor- is the mass of the crosshead hardware in that Wartsila engine counted as part of the 10,000lb piston weight?

I also noted that the 300 ton crankshaft in that engine does not use any counterweighting. So I would not imagine that piston/conrod inertia loads are a problem with respect to the main bearings.