Engineering Connecting Rod Question - LS1GTO.com Forums

Vendors


Go Back   LS1GTO.com Forums > GTO Tech > LS1/LS2 Engine Tech

Reply
 
Thread Tools Rate Thread
Old 08-29-2013, 08:26 PM   #1
FSAE_Junkie
Registered User
 
FSAE_Junkie's Avatar
 
2006 GTO Owner

Join Date: Mar 2008
Location: Louisville/Cincinnati
Posts: 574


Offline
Engineering Connecting Rod Question

So, about a year ago, I bought a 4-axis CNC in the hopes of doing something with it. At the time, one of my thoughts was that I could easily make some billet connecting rods for fun and extremely cheap. So, I set off to do some preliminary loads estimation, build some finite element models, look at fatigue life, and run some of my fracture mechanics/probabilistic lifing codes on it.

Well, I ended up shelving the project as there were some more pressing GTO projects to get done, but something bothered me about the loads calculations I had done.

Everybody talks about RPM killing rods, but I just don't see how that's possible ( I'm not talking bearing ovalization, but rod fracture). Specifically, the thought that TDC of the exhaust stroke is what kills rods. But let's take a look at your typical 3.622" stroke, 8,000 RPM LS engine.

- At TDC, the piston (and we'll assume the top half of the rod) have 3,300 g's acting on them.
- At 1.225 lbm for the piston and top half of the rod, that's 5300 lbf pulling the rod apart.


Now let's look at the compression stroke on a 1200 bhp (~950 rwhp) motor. From some prior combustion simulation I ran on a specific combination:
- Peak cylinder pressures are on the order of 2500 psi.
- Across 12.5 in^2 of piston surface area, that's about 31,000 lbf in compression.
- Even subtracting the piston loads near TDC, you still have 26,000 lbf wanting to buckle your rod.

Any thoughts on why people like to say RPM destroys rods when in a medium power setup, it appears the tensile stress is only about 1/5 of the compressive?

I'm legitimately just looking for some discussion - I can't help but think I'm missing something.
__________________
GTO#2: 2005 Manual 4l80e - Forged LS2 - Ported 243's - F1A - E85

GTO #1: 2006 M6 - SOLD
Cam/LT's/Slicks
440 rwhp/407 rwtq
11.95 @ 115.1 - 1.68 60'

Last edited by FSAE_Junkie; 08-29-2013 at 08:37 PM.
  Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Old 08-29-2013, 10:58 PM   #2
Raymond Luxury Yacht
Suffering Fools
 
Raymond Luxury Yacht's Avatar
 

Join Date: Apr 2005
Location: Manvel, TX
Posts: 15,237


Offline
Does the metal react differenty under tension vs compression?
__________________
Quote:
Originally Posted by Aristotle...View Post
Humor is the only test of gravity, and gravity of humor; for a subject which will not bear raillery is suspicious, and a jest which will not bear serious examination is false wit. -- Aristotle

Nassau Blue - 407/390 http://www.fquick.com/Lothar-[Tech Spec]
  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:04 AM   #3
elephantrider
pea on mea
 
elephantrider's Avatar
 
2004 GTO Owner

Join Date: Apr 2012
Location: round rock texas
Posts: 4,016


Offline
in for smarts.
__________________
trico wiper blades
  Reply With Quote Quick reply to this message
 
Old 08-30-2013, 04:24 AM   #4
Steel Chicken
Cthulhu's IT D1r3ktФя
 
Steel Chicken's Avatar
 
2004 GTO Owner

Join Date: Dec 2004
Location: Castle Rock, CO
Posts: 20,246
Premium Member


Offline
I found this very helpful in causing my head to explode.

http://www.eng.utoledo.edu/mime/facu...Cpp615-624.pdf
__________________
2004 GTO, Stock LS1 internals, EDC custom cam, TVS2300 @ 14PSI, ETP heads and corn juice. 742 RAWRZORS.

Quote:
Justice is swift and unfair in the eyes of those who don't know shit. - tmiller

  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:29 AM   #5
elephantrider
pea on mea
 
elephantrider's Avatar
 
2004 GTO Owner

Join Date: Apr 2012
Location: round rock texas
Posts: 4,016


Offline
mmm, yes... that attachment should keep the op busy
for a bit.

In order to capture the structural behaviour of the
connecting rod under service operating conditions,
quasi-dynamic FEA was performed. An FE model
mesh with about 105 parabolic tetrahedral elements,
with uniform global element length of 1.5 mm and
local element length of 1 mm at locations with
chamfers, was used. As a connecting rod is
designed for very long life, stresses are in the elastic
range, and as a result linear elastic analysis was
conducted.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:35 AM   #6
Steel Chicken
Cthulhu's IT D1r3ktФя
 
Steel Chicken's Avatar
 
2004 GTO Owner

Join Date: Dec 2004
Location: Castle Rock, CO
Posts: 20,246
Premium Member


Offline
Quote:
Originally Posted by elephantrider...View Post
mmm, yes... that attachment should keep the op busy
for a bit.

In order to capture the structural behaviour of the
connecting rod under service operating conditions,
quasi-dynamic FEA was performed. An FE model
mesh with about 105 parabolic tetrahedral elements,
with uniform global element length of 1.5 mm and
local element length of 1 mm at locations with
chamfers, was used. As a connecting rod is
designed for very long life, stresses are in the elastic
range, and as a result linear elastic analysis was
conducted.

Translation, the vector of the stresses on the con rod move around as the engine performs a revolution, drastically altering both the force of the loads and location off the stress on the con rod itself.

So, if you just look at compression pushing down on top of the cylinder, you are not seeing the whole picture.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:51 AM   #7
mistermike
May I quote you on that?
 
mistermike's Avatar
 
2004 GTO Owner

Join Date: Dec 2003
Location: Cleveland
Posts: 22,049
GKPW52

Offline
I've heard that RPM stresses rod bolts, but not the rods themselves. The idea that high power stresses rod bolts is probably just as wrong as saying RPM kills the rods.
__________________
http://i67.photobucket.com/albums/h306/MikeStainlessWorks/SigPicAniSm.gif
World's First Twin Screw Pontiac GTO

Blown H/C 408 and other stuff. Billet tires. Plasti Dip tints. Stage 3 keyboard. /////.
2017 Chevy SS w/ dead hooker in spacious trunk
2017 Cruze diesel A9 MPG queen. Rollin coal w / my balls in spacious glove compartment.
fquick
  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:54 AM   #8
05ls7gto
Original Owners Club
 
05ls7gto's Avatar
 
2006 GTO Owner

Join Date: Nov 2006
Location: Cape Coral, FL
Posts: 9,471


Offline
Quote:
Originally Posted by Steel Chicken...View Post
I found this very helpful in causing my head to explode.

http://www.eng.utoledo.edu/mime/facu...Cpp615-624.pdf

Awesome article, thanks for sharing.
__________________
R.I.P. Renold K Algoo 9/4/1986 - 11/29/2006

2005 QSM/Anthracite/17/A4 - R.I.P.
2005 MBM/Anthracite/17/A4 - Mod List
2006 SRM/Anthracite/17/M6 - Mod List
2014 PBM Chevy SS/Sunroof/No Spare
2016 SLIHRM Chevy SS/M6/Sunroof/Spare
  Reply With Quote Quick reply to this message
Old 08-30-2013, 04:59 AM   #9
dcarino_04
Registered User
 

Join Date: Jan 2011
Location: illinois
Posts: 209


Offline
Different materials have different strengths. Metals seem to hold up decent on both compression and tension. However our bones suck at tension but exceed by far in compression. I think you would be correct saying it's different but I'm not sure what billet 's attributes are

Sent from AutoGuide.com Free App
  Reply With Quote Quick reply to this message
Old 08-30-2013, 05:00 AM   #10
Steel Chicken
Cthulhu's IT D1r3ktФя
 
Steel Chicken's Avatar
 
2004 GTO Owner

Join Date: Dec 2004
Location: Castle Rock, CO
Posts: 20,246
Premium Member


Offline
Quote:
Originally Posted by mistermike...View Post
I've heard that RPM stresses rod bolts, but not the rods themselves. The idea that high power stresses rod bolts is probably just as wrong as saying RPM kills the rods.

Why would RPM not be a major factor in killing the rods?

Loads increase with the SQUARE of engine RPM.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 05:02 AM   #11
dcarino_04
Registered User
 

Join Date: Jan 2011
Location: illinois
Posts: 209


Offline
Seeing this makes me look forward to mechanics of materials next semester lol

Sent from AutoGuide.com Free App
  Reply With Quote Quick reply to this message
Old 08-30-2013, 05:12 AM   #12
The Great FPFQ of 2013
Unaware of loss, Nor of gain
 
The Great FPFQ of 2013's Avatar
 
2006 GTO Owner

Join Date: Jul 2013
Location: New Orleans
Posts: 7,712


Offline
Quote:
Originally Posted by mistermike...View Post
I've heard that RPM stresses rod bolts, but not the rods themselves. The idea that high power stresses rod bolts is probably just as wrong as saying RPM kills the rods.



I guess his hands should have had arp hardware.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 05:24 AM   #13
LS2-GTO
1 of 40,757
 
LS2-GTO's Avatar
 
2006 GTO Owner

Join Date: Aug 2006
Location: New York
Posts: 6,106
Premium Member


Offline
Bearings are typically the weak point.
Bent rods are often seized at the crank.
__________________
  Reply With Quote Quick reply to this message
Old 08-30-2013, 05:51 AM   #14
mistermike
May I quote you on that?
 
mistermike's Avatar
 
2004 GTO Owner

Join Date: Dec 2003
Location: Cleveland
Posts: 22,049
GKPW52

Offline
My rod was seized at the border.
WMD, baby.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 06:32 AM   #15
FSAE_Junkie
Registered User
 
FSAE_Junkie's Avatar
 
2006 GTO Owner

Join Date: Mar 2008
Location: Louisville/Cincinnati
Posts: 574


Offline
I'm glad this thread got some hits. I'm at work now, so I can't post up too much at the moment, but I will make a few comments that I was going to make last night, but decided to wait on to not clutter up the post.

Steel, I'll have to take a look at that link tonight.


From a high cycle fatigue calculation, you can normalize a 'tension-compression' cycle into a "0 - max - 0" stress cycle using both the positive stress and negative stress - implying that both are important.

However, there are differences in tension and compression in metals. From a fracture mechanics point of view, only tensile forces serve to start and propagate cracks which leads to premature failure. However, this is very closely related to fatigue failure, and as mentioned above, fatigue depends on both tension and compression. I've never been able to get clarification on this.

Mistermike, that was another point I was going to make. Rating a rod bolt in terms of hp/tq makes very little sense to me. They only see the tensile stress due to RPM.

Also, the above quote from the article mentioned some rod FEA. I've got some simple FEA analysis I can share when I get home.

LS2-GTO hits on the only thing I could think of. Excessive RPM will ovalize the rod bearing which could lead to a spun bearing. At that point, all sorts of things could happen to the rod which might make it appear that the rod itself failed.

I brought this topic up because I've got an engine in the machine shop and I'm half tempted to want to run this thing up to 8,000 RPM's. This got me looking at what is required of the rod. I'm coming to the conclusion that the rod journal is more critical than the rod itself from an RPM standpoint (I could very well be wrong, this is just my observation). I've read of builders who hone the rods with a sideways clamping pressure to make the 3:00 and 9:00 position of the journal about 0.0005" wider. So at high RPM's when it does start to ovalize, it doesn't clamp down on the journal.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 06:43 AM   #16
Steel Chicken
Cthulhu's IT D1r3ktФя
 
Steel Chicken's Avatar
 
2004 GTO Owner

Join Date: Dec 2004
Location: Castle Rock, CO
Posts: 20,246
Premium Member


Offline
Quote:
Originally Posted by FSAE_Junkie...View Post
Mistermike, that was another point I was going to make. Rating a rod bolt in terms of hp/tq makes very little sense to me. They only see the tensile stress due to RPM.

Tensile stress is something most people won't know what to do with. And while tensile stress is very vague with regards to application, HP ratings (if used for a specific application are reasonably accurate) are something people can understand. A rod bolt that handles x amount of tensile stress on a testing machine will result in DIFFERENT HP/TQ values in different motors depending on stroke and other factors. So in that case, rating by HP/TQ BY APPLICATION actually makes more sense.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 09:28 AM   #17
Kanding
Registered User
 
Kanding's Avatar
 
2006 GTO Owner

Join Date: May 2006
Location: Michigan
Posts: 113


Offline
FSAE,

Out of curiosity, what FEA code are you using, and what type of analysis do you plan to do?

I didn’t read through Shenoy’s paper thoroughly, but they conclude that a static analysis would provide unrealistic stress results and a dynamic analysis is needed (but here they didn’t really model the kinematics of the rod, but more simply applied time-varying loads to a static mesh—I assume this is what they meant by calling their analysis ‘quasi-dynamic’).
__________________
Kanding
CGM 06 M6 18" stock 1 of 466
  Reply With Quote Quick reply to this message
Old 08-30-2013, 10:08 AM   #18
Shawshank
Cyclone Gray Moneypit
 
Shawshank's Avatar
 
2005 GTO Owner

Join Date: Oct 2012
Location: Texas
Posts: 1,989


Offline
Quote:
Originally Posted by Steel Chicken...View Post
Tensile stress is something most people won't know what to do with. And while tensile stress is very vague with regards to application, HP ratings (if used for a specific application are reasonably accurate) are something people can understand. A rod bolt that handles x amount of tensile stress on a testing machine will result in DIFFERENT HP/TQ values in different motors depending on stroke and other factors. So in that case, rating by HP/TQ BY APPLICATION actually makes more sense.

This would actually be an interesting rating to give components. It seems that with aftermarket rotating assemblies (such as forged) the rod bolts are generally the rated component since the forging has such a higher strength proportionally (making the bolts the weak point).

It would be cool to see ratings for a given assembly and hardware based on design limitation by RPM and application like you mentioned! This may further illuminate and distinguish the quality of different aftermarket manufacturers.

Quote:
Originally Posted by Kanding...View Post
FSAE,

Out of curiosity, what FEA code are you using, and what type of analysis do you plan to do?

I didn’t read through Shenoy’s paper thoroughly, but they conclude that a static analysis would provide unrealistic stress results and a dynamic analysis is needed (but here they didn’t really model the kinematics of the rod, but more simply applied time-varying loads to a static mesh—I assume this is what they meant by calling their analysis ‘quasi-dynamic’).

I believe you hit the nail on the head. This is exactly what I would assume 'quasi-dynamic' to mean; applying the changing loads during rotation to the same mesh.

If they were able to correlate the 'quasi-dynamic' model to a true dynamic analysis, it would be interesting to see if this could be tested in a real application. It may be a simplified representation to design a fixture for testing the rotating assembly, but it could provide direction for determining more accurately the true stresses. A fully assembled engine would probably be ideal, but I'm not entirely sure how one could monitor the stresses/strains in the rotating assembly (strain gauges with memory or means of transmitting data real-time?). Either way, they would need to pretty robust to survive in that sort of environment!
  Reply With Quote Quick reply to this message
Old 08-30-2013, 10:23 AM   #19
machinistone
Registered User
 
machinistone's Avatar
 
2006 GTO Owner

Join Date: Jul 2006
Location: Santa Rosa, CA
Posts: 1,633


Offline
Quote:
Originally Posted by mistermike...View Post
I've heard that RPM stresses rod bolts, but not the rods themselves. The idea that high power stresses rod bolts is probably just as wrong as saying RPM kills the rods.

Yep.

Quote:
Originally Posted by FSAE_Junkie...View Post
I'm glad this thread got some hits. I'm at work now, so I can't post up too much at the moment, but I will make a few comments that I was going to make last night, but decided to wait on to not clutter up the post.

Steel, I'll have to take a look at that link tonight.


From a high cycle fatigue calculation, you can normalize a 'tension-compression' cycle into a "0 - max - 0" stress cycle using both the positive stress and negative stress - implying that both are important.

However, there are differences in tension and compression in metals. From a fracture mechanics point of view, only tensile forces serve to start and propagate cracks which leads to premature failure. However, this is very closely related to fatigue failure, and as mentioned above, fatigue depends on both tension and compression. I've never been able to get clarification on this.

Mistermike, that was another point I was going to make. Rating a rod bolt in terms of hp/tq makes very little sense to me. They only see the tensile stress due to RPM.

Also, the above quote from the article mentioned some rod FEA. I've got some simple FEA analysis I can share when I get home.

LS2-GTO hits on the only thing I could think of. Excessive RPM will ovalize the rod bearing which could lead to a spun bearing. At that point, all sorts of things could happen to the rod which might make it appear that the rod itself failed.

I brought this topic up because I've got an engine in the machine shop and I'm half tempted to want to run this thing up to 8,000 RPM's. This got me looking at what is required of the rod. I'm coming to the conclusion that the rod journal is more critical than the rod itself from an RPM standpoint (I could very well be wrong, this is just my observation). I've read of builders who hone the rods with a sideways clamping pressure to make the 3:00 and 9:00 position of the journal about 0.0005" wider. So at high RPM's when it does start to ovalize, it doesn't clamp down on the journal.

I've only done the clamp trick with motors that have to use an OEM rod that doesn't have the rigidity of stronger aftermarket ones. Also use a higher eccentricity P style bearing instead of an H so it's less likely to pinch the oil film at the parting line of the rod. This is why we strive to use the lightest pistons possible in high rpm applications, because the load that the big end of the rod and the bolts see due to the acceleration and deceleration of the piston is what causes rods to break much more often than pure compression load from how much power the engine makes.

Here's a little video of a set of titanium motorcycle rods my buddy around the corner designed and made.

http://web.archive.org/web/201102021...pe.com/rad.swf
__________________
2011 335i M Sport
'06 GTO CGM M6 (sold)
'66 Mustang Fastback
  Reply With Quote Quick reply to this message
Old 08-30-2013, 11:52 AM   #20
Lakegoat
Rambler is stuck in 2nd gear
 
Lakegoat's Avatar
 
2005 GTO Owner

Join Date: Apr 2007
Location: Florida
Posts: 2,055


Offline
Looks like it is humping a dry hole.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 12:41 PM   #21
Kanding
Registered User
 
Kanding's Avatar
 
2006 GTO Owner

Join Date: May 2006
Location: Michigan
Posts: 113


Offline
Quote:
Originally Posted by Shawshank...View Post
If they were able to correlate the 'quasi-dynamic' model to a true dynamic analysis, it would be interesting to see if this could be tested in a real application. It may be a simplified representation to design a fixture for testing the rotating assembly, but it could provide direction for determining more accurately the true stresses. A fully assembled engine would probably be ideal, but I'm not entirely sure how one could monitor the stresses/strains in the rotating assembly (strain gauges with memory or means of transmitting data real-time?). Either way, they would need to pretty robust to survive in that sort of environment!

I’ve never done engine modeling before, but I would have thought these fairly fundamental modeling issues would have been resolved some time ago, as car motors have existed so long.

I know at least two general purpose commercial FEA codes that can be used to model the kinematics as well as the kinetics of the motor, but of course building the model is more difficult (and error prone), and time consuming to run, than what was done in the paper.

You raise a good point with experimental verification. All of the gages that I used before were wired and fairly easily damaged—not practical on a working rotating assembly. However, I guess it even surprises me that this wasn’t somehow resolved before as well, given the age and significance of the problem.
  Reply With Quote Quick reply to this message
Old 08-30-2013, 06:45 PM   #22
Oh4GTO
dum dum dum dum dudum
 
Oh4GTO's Avatar
 
2004 GTO Owner

Join Date: Jun 2005
Location: Atlanta
Posts: 6,217


Offline
Send a message via AIM to Oh4GTO
^^^ Im sure F1 teams have the means to test all of this considering the ridiculous motors they build.

Going back to OP's question. Modeling the strengths and stressors of a static piece or in a theoretically set environment where each parameter is the same is one thing. When you throw an engine into a vehicle that has varying temperatures throughout the block, inconsistent air and fuel, different loads and actually moving a few thousand pounds, anything can happen. Plus you have inconsistencies in the manufacturing of the actual product.

My educated guess is that Eagle, Comp....etc. take their products and test them for real and see where the failures occur. Subtract a few hundred horsepower and rate their product for xxx HP leaving plenty of buffer.
__________________

LS3 longblock, LS9 cam, LS9 headgaskets, NW 90mm, APS 5psi = 610 RWHP / 578 RWTQ
  Reply With Quote Quick reply to this message
Old 08-30-2013, 06:52 PM   #23
Shawshank
Cyclone Gray Moneypit
 
Shawshank's Avatar
 
2005 GTO Owner

Join Date: Oct 2012
Location: Texas
Posts: 1,989


Offline
Quote:
Originally Posted by Oh4GTO...View Post
^^^ Im sure F1 teams have the means to test all of this considering the ridiculous motors they build.

Going back to OP's question. Modeling the strengths and stressors of a static piece or in a theoretically set environment where each parameter is the same is one thing. When you throw an engine into a vehicle that has varying temperatures throughout the block, inconsistent air and fuel, different loads and actually moving a few thousand pounds, anything can happen. Plus you have inconsistencies in the manufacturing of the actual product.

My educated guess is that Eagle, Comp....etc. take their products and test them for real and see where the failures occur. Subtract a few hundred horsepower and rate their product for xxx HP leaving plenty of buffer.

I'd agree with your analysis. It would probably take much less time and funding to put together a long block and run it at elevated RPM for millions of cycles and tear it down at periodic intervals and performing some form of dull grading. Perhaps there is some form of endurance test as well with a teardown at the end, but these are things I'd probably try to do haha.

I'm sure there are some decent models out there that GM would use to predict wear and correlate to actual testing. I'm just not sure to what extent aftermarket suppliers would go in order to test their product. I feel they'd be a bit more likely to overbuild it, slap on a fatty safety factor, and rate it based on worst case scenario.

Sub'd for moar knowledge
  Reply With Quote Quick reply to this message
Old 08-30-2013, 07:47 PM   #24
Oh4GTO
dum dum dum dum dudum
 
Oh4GTO's Avatar
 
2004 GTO Owner

Join Date: Jun 2005
Location: Atlanta
Posts: 6,217


Offline
Send a message via AIM to Oh4GTO
I used to live in Daytona Beach and there was a race team out of there called Spirit of Daytona. Im sure some of you remember when they ran 2 GTOs in the Koni Challenge. They also ran to Dayton Prototype cars. They were both GM backed so they ran a forged LS motor. The few times that the engine failed was due to a rod coming apart and putting a breather window in the side of the block.

Now that I think about it, a lot of failures that I have seen or read about occurred during a gear change. This is probably more violent than a consistent high RPM setting. With a gear change you have the rod traveling at its peak speed at xxxx RPM and instantly put a huge load on it while still demanding maximum power from the engine. It is this change that more than likely destroys the rod.
  Reply With Quote Quick reply to this message
Old 08-31-2013, 05:33 AM   #25
Mad Martigan
Rock Lobster!
 
Mad Martigan's Avatar
 
2005 GTO Owner

Join Date: Jun 2013
Location: Tucker, GA
Posts: 97


Offline
Quote:
Originally Posted by FSAE_Junkie...View Post

I brought this topic up because I've got an engine in the machine shop and I'm half tempted to want to run this thing up to 8,000 RPM's. This got me looking at what is required of the rod. I'm coming to the conclusion that the rod journal is more critical than the rod itself from an RPM standpoint (I could very well be wrong, this is just my observation). I've read of builders who hone the rods with a sideways clamping pressure to make the 3:00 and 9:00 position of the journal about 0.0005" wider. So at high RPM's when it does start to ovalize, it doesn't clamp down on the journal.

Might be a low brow question considering the content of the thread so far, but how does honing rods that way affect long term wear? Or is it purely for racing applications with relatively frequent rebuilds?
__________________
ROCK LOBSTER!!!!??!?!?!?!?!
  Reply With Quote Quick reply to this message
Old 08-31-2013, 09:26 AM   #26
LSPerformance
Registered User
 
2004 GTO Owner

Join Date: Oct 2006
Location: Charlotte, NC
Posts: 565


Offline
With RPM increased beyond the rods capability the big end bores will pinch in at the parting lines, cutting off the oil wedge.

I used to clamp/pinch the rods at the parting line smaller by .0025" before honing to size on OE FE Ford rods used for Super Stock drag racing. Release the clamp, and the bore has .0025" eccentricity at the parting line.

My last LS build had Manley H Lites that were pinching in at the parting lines. I could see the evidence in looking at the bearings. I ran them to 8300 rpm, so I likely went beyond what that rod's big end geometry was capable of withstanding. My new 8500+ rpm engine build is getting Carrillo rods.
  Reply With Quote Quick reply to this message
Old 09-08-2013, 09:14 AM   #27
FSAE_Junkie
Registered User
 
FSAE_Junkie's Avatar
 
2006 GTO Owner

Join Date: Mar 2008
Location: Louisville/Cincinnati
Posts: 574


Offline
Quote:
Originally Posted by Steel Chicken...View Post
Why would RPM not be a major factor in killing the rods?

Loads increase with the SQUARE of engine RPM.

I think the main point is that at stock power levels, the RPM based stress is equal to if not more than stress from combustion pressure. For example, in that paper you referenced, their combustion pressure was only 35 bar and so the RPM based stress at 5700 RPM was higher than due to combustion. But this was a lower power motor in OEM configuration.

Take a stock LS motor that can rev to 6500 RPM easily. If you increase the redline to 8000 RPM, the stress will go up by a factor of 8000^2/6500^2. This is a 50% increase. This is certainly a large increase, but...

If you go from a stock engine with 50 bar combustion pressure (400 hp) to one with 150 bar combustion pressure (1200 hp), you've increased the compressive load by 300%...

So, it's interesting to me that people will build a boosted, 900 rwhp motor, but then worry about spinning it 500 RPM higher (excluding valvetrain considerations).


Quote:
Originally Posted by Steel Chicken...View Post
Tensile stress is something most people won't know what to do with. And while tensile stress is very vague with regards to application, HP ratings (if used for a specific application are reasonably accurate) are something people can understand. A rod bolt that handles x amount of tensile stress on a testing machine will result in DIFFERENT HP/TQ values in different motors depending on stroke and other factors. So in that case, rating by HP/TQ BY APPLICATION actually makes more sense.

What I was trying to get at is that if you look at where the forces go, the rods see no stress from the power stroke. The wrist pin is being forced downwards towards the rod journal and the rod is in compression. The bolt won't see any of this load. Bolt's only see stress due to RPM. I'm pretty well convinced you could run stock GM rod bolts in a 2000 hp combo so long as your RPM was not much higher than stock.

Though I agree, trying to rate them by RPM would be difficult as it would require a table of RPM and stroke, and piston weight, etc... And really, manufacturers are just covering their butts by keeping things vague. Like forged I-beams that they rate at 650 bhp, while a stock LS2 rod has survived 1200 bhp (albeit not for long). We all know the forged rod is severly underrated. From a liability perspective, I'm sure they don't want to commit to saying, "yes, our rod will handle 10,430 RPM in an LS motor with 620 g piston and 3.622" stroke".

Quote:
Originally Posted by Kanding...View Post
FSAE,

Out of curiosity, what FEA code are you using, and what type of analysis do you plan to do?

I didn’t read through Shenoy’s paper thoroughly, but they conclude that a static analysis would provide unrealistic stress results and a dynamic analysis is needed (but here they didn’t really model the kinematics of the rod, but more simply applied time-varying loads to a static mesh—I assume this is what they meant by calling their analysis ‘quasi-dynamic’).

So, I have only run what they would call a static analysis. I still apply a restraint to the rod. I do not apply loads only, and then attempt to stop the rigid body motion that ensues. Though after reading the paper, I'm still not sure how their quasi-dynamic analysis differed from a static analysis. They still manually applied the accelerations to the pin and crank ends.

From a personal engineering perspective, Fatigue is an enormously complicated subject. There is no explicit textbook answer. You really need fatigue data specifically for the material you are using and even then, you need to know how to correlate your FE model results (tensile and compressive stress) to the curves, because very rarely do you test your specimen to the exact stress range it will see in operation. Instead, you usually test from say -60 ksi to +60 ksi and then use curve fits to adjust it to a 0 -max - 0 cycle. Additionally, were the fatigue tests strain controlled or load controlled? As the part gets nearer to failure, and a crack has propagated across half your specimen, the type of loading makes a huge difference on the life you will measure.

On top of this, to correctly life a component, you need to know it's 'mission'. You don't want to design your rod for 10 million cycles at max RPM and power because it will be way over designed. Additionally, you can't ignore the loads under cruise conditions. So you have to simulate a mix of conditions and then use a cumulative damage theory to estimate life.

My point of all this being that there are a lot of assumptions that have to be made even after calculating a stress - and in my opinion, these assumptions have a much larger 'variability' than you'd see from small differences in modeling assumptions.


Quote:
Originally Posted by Shawshank...View Post
This would actually be an interesting rating to give components. It seems that with aftermarket rotating assemblies (such as forged) the rod bolts are generally the rated component since the forging has such a higher strength proportionally (making the bolts the weak point).

It would be cool to see ratings for a given assembly and hardware based on design limitation by RPM and application like you mentioned! This may further illuminate and distinguish the quality of different aftermarket manufacturers.

I believe you hit the nail on the head. This is exactly what I would assume 'quasi-dynamic' to mean; applying the changing loads during rotation to the same mesh.

If they were able to correlate the 'quasi-dynamic' model to a true dynamic analysis, it would be interesting to see if this could be tested in a real application. It may be a simplified representation to design a fixture for testing the rotating assembly, but it could provide direction for determining more accurately the true stresses. A fully assembled engine would probably be ideal, but I'm not entirely sure how one could monitor the stresses/strains in the rotating assembly (strain gauges with memory or means of transmitting data real-time?). Either way, they would need to pretty robust to survive in that sort of environment!

People have done remote piston temp measurement by using thermocouples with microwave transmitters. I would think something similar could be done with a rod and strain gage. However, with that said, I'm not sure it would buy you much they probably don't already know. I am confident that the OEM's have their own proprietary tools specifically for analyzing connecting rods. It would not surprise me if they've done this at one point to validate their model, but now use the model near exclusively.

Quote:
Originally Posted by machinistone...View Post
Yep.

I've only done the clamp trick with motors that have to use an OEM rod that doesn't have the rigidity of stronger aftermarket ones. Also use a higher eccentricity P style bearing instead of an H so it's less likely to pinch the oil film at the parting line of the rod. This is why we strive to use the lightest pistons possible in high rpm applications, because the load that the big end of the rod and the bolts see due to the acceleration and deceleration of the piston is what causes rods to break much more often than pure compression load from how much power the engine makes.

Here's a little video of a set of titanium motorcycle rods my buddy around the corner designed and made.

http://web.archive.org/web/201102021...pe.com/rad.swf

That makes a lot of sense.

Did you buddy end up running the Ti rods? My biggest hold up to making rods myself is all the specialized steps needed. I've been told my CNC doesn't have fast enough Z travel to hone the rods correctly (and get a cross-hatch pattern, though I'm still not sure why this is important since the bearing doesn't move like pistons do in a cylinder). So there's an added expensive to taking them to a shop for final sizing. Then having to make your own bronze bushings, pressing them in, and taking them to final size. Lastly, a high hp rod should really be shot peened. Oh, and you also have to make sure you get the heat treat right.

I've looked into Ti, but I can't the material for near cheap enough considering how much get's wasted from a piece of billet.

Quote:
Originally Posted by Kanding...View Post
I’ve never done engine modeling before, but I would have thought these fairly fundamental modeling issues would have been resolved some time ago, as car motors have existed so long.

I know at least two general purpose commercial FEA codes that can be used to model the kinematics as well as the kinetics of the motor, but of course building the model is more difficult (and error prone), and time consuming to run, than what was done in the paper.

You raise a good point with experimental verification. All of the gages that I used before were wired and fairly easily damaged—not practical on a working rotating assembly. However, I guess it even surprises me that this wasn’t somehow resolved before as well, given the age and significance of the problem.

What codes are you thinking of?

Quote:
Originally Posted by Oh4GTO...View Post
^^^ Im sure F1 teams have the means to test all of this considering the ridiculous motors they build.

Going back to OP's question. Modeling the strengths and stressors of a static piece or in a theoretically set environment where each parameter is the same is one thing. When you throw an engine into a vehicle that has varying temperatures throughout the block, inconsistent air and fuel, different loads and actually moving a few thousand pounds, anything can happen. Plus you have inconsistencies in the manufacturing of the actual product.

My educated guess is that Eagle, Comp....etc. take their products and test them for real and see where the failures occur. Subtract a few hundred horsepower and rate their product for xxx HP leaving plenty of buffer.

I don't think it would just be F1, I'm sure any of the OEM's have a way to test this. As for the modeling, the variability is the fun part

Quote:
Originally Posted by Shawshank...View Post
I'd agree with your analysis. It would probably take much less time and funding to put together a long block and run it at elevated RPM for millions of cycles and tear it down at periodic intervals and performing some form of dull grading. Perhaps there is some form of endurance test as well with a teardown at the end, but these are things I'd probably try to do haha.

I'm sure there are some decent models out there that GM would use to predict wear and correlate to actual testing. I'm just not sure to what extent aftermarket suppliers would go in order to test their product. I feel they'd be a bit more likely to overbuild it, slap on a fatty safety factor, and rate it based on worst case scenario.

I would agree that the aftermarket likely doesn't waste a lot of time/money for testing. When you're not worried about removing $0.25 from each rod for 10 million rods, it's easy enough to build something that is much more robust than the OEM produces.


Quote:
Originally Posted by Mad Martigan...View Post
Might be a low brow question considering the content of the thread so far, but how does honing rods that way affect long term wear? Or is it purely for racing applications with relatively frequent rebuilds?

I would be curious to know the answer to this myself.
  Reply With Quote Quick reply to this message
Old 09-08-2013, 10:01 AM   #28
FSAE_Junkie
Registered User
 
FSAE_Junkie's Avatar
 
2006 GTO Owner

Join Date: Mar 2008
Location: Louisville/Cincinnati
Posts: 574


Offline
As promised, here's some output from some quick FEA I had run. This is a pretty simple model as it does not include bolts, or their pre-tension loads. This was just a 1st step in sizing the rod for the conditions I would like to run.

Then peak stress points for the tension case will be easily remedied using more material in the top of the pin end. Confident I could drop the stress down to 30 ksi.

The compressive load however is quite high. At Rc 42 hardness, 4340 will have a yield strength near 180,000 psi. However, it's 10 million cycle endurance limit will be down around 45,000 psi. This goes back to my point earlier about not being able to design for a full endurance life under maximum operating conditions. I'd like to get the compressive stress down to no higher than 60 ksi.

  Reply With Quote Quick reply to this message
Old 09-08-2013, 06:32 PM   #29
Kanding
Registered User
 
Kanding's Avatar
 
2006 GTO Owner

Join Date: May 2006
Location: Michigan
Posts: 113


Offline
Very cool and interesting analysis, what code did you use?

Quote:
Originally Posted by FSAE_Junkie...View Post
What codes are you thinking of?

I’m a structural engineer not a ME, so I don’t know any automotive/mechanical specific codes, but LS-DYNA and ELS can both model large displacement, transient dynamic, rigid body motion. I’ve used both to model very high displacement problems with element contact (primarily blast analysis with element separation and car crash analysis). Both use explicit solvers, however, so solution convergence is not guaranteed. Maybe ABAQUS too, it has a decent large displacement solver, but I’m not sure it can model displacement as large as a rotating assembly.


Quote:
Originally Posted by FSAE_Junkie...View Post
Though after reading the paper, I'm still not sure how their quasi-dynamic analysis differed from a static analysis.

Me either, other than there was some mention of time-varying loads. But, dynamic to me implies that mass has an impact on the analysis (i.e. solving ma+cv+ku=F(t)), but I didn’t read the paper carefully enough to see if inertial effects were really accounted for.


Thanks for posting this, very nice.
  Reply With Quote Quick reply to this message
Old 09-08-2013, 07:54 PM   #30
machinistone
Registered User
 
machinistone's Avatar
 
2006 GTO Owner

Join Date: Jul 2006
Location: Santa Rosa, CA
Posts: 1,633


Offline
Quote:
Originally Posted by FSAE_Junkie...View Post
That makes a lot of sense.

Did you buddy end up running the Ti rods? My biggest hold up to making rods myself is all the specialized steps needed. I've been told my CNC doesn't have fast enough Z travel to hone the rods correctly (and get a cross-hatch pattern, though I'm still not sure why this is important since the bearing doesn't move like pistons do in a cylinder). So there's an added expensive to taking them to a shop for final sizing. Then having to make your own bronze bushings, pressing them in, and taking them to final size. Lastly, a high hp rod should really be shot peened. Oh, and you also have to make sure you get the heat treat right.

I've looked into Ti, but I can't the material for near cheap enough considering how much get's wasted from a piece of billet.

Yes, they designed and built several different sets over the years. Their Ti rods were not heat treated nor shot peened, what ever hardness the bar stock came with is what they got. For fully machined steel rods you would want to have them shot peened, and it would have to be done properly, having the local machine shop run them through their shot blasting cleaner cabinet is not even close to the same as an approved shot peen process. I have seen rods fail from improper shot peening procedures.

They do not need to have a honed cross hatch finish on the id as long as you can hold the dimensional tolerance to .0002, I would guess that the Ra requirement is < 40 which any decent mill should be able to accomplish with the right tooling.

The bronze bushings are available from many difference sources in varying ODs, you would simply press them in and bore on center and could use a ball hone to put in a cross hatch for oil retention if you can again hold that bore tolerance to .0002"

If you cannot meet that dimensional tolerance then you would want to finish bore the big and small end to within .002" or so of final size and then have a machine shop hone them to size which shouldn't cost more than $100 for a set of eight when all they need to do is finish hone.
  Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the LS1GTO.com Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:
Location
Where you live (must be accurate to IP location, if wrong you may be permanently banned)
Insurance
Please select your insurance company (Optional)

Log-in


Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -8. The time now is 06:52 PM.




Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
vBulletin Security provided by vBSecurity v2.2.2 (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.