DIVERGENT THERMAL FEEDBACK LOOPS IN TURBOCHARGER APPLICATIONS—GM CASE STUDY
What if, when a bullet was fired, it didn’t slow down, but rather sped up the further away it got? Imagine the insanity, if every time you drank water, you became more thirsty, or if a candle burned faster the further down it got. These analogies resemble a divergent mechanism that exists in the cooling system of some turbocharged vehicles. In order for a vehicles cooling system to exhibit stable and effective operation over a wide range of conditions, conditions must allow the system to seek thermal equilibrium at a set point temperature…flow thermostats regulate this. The addition of after-cooled turbochargers complicates the cooling system, with a heat producing charge air cooler, CAC, unfortunately located in front of the radiator. While this article focuses on the LLY changes, it also applies to every after cooled vehicle in the world. The GM produced LLY Duramax diesel motor is a prominent example which suffers complexities leading to cooling system failures. This detailed article chronicles the mystery to one of GM’s biggest cooling system debacles in modern vehicle production history, a complex time lapsed recipe over an evolved “countdown to disaster”.
Authors Note: This is a technical paper that relies on a fundamental understanding of fluid flow and heat transfer, and to a lesser extent, thermodynamics. For ease of translation, much of the technical theory is omitted. But it would be useful for the nontechnical reader to accept a real concept: that when air flows over a surface or in a tube, friction results from its contact with the surface. It is no different than trying to push a lot of air through a very narrow straw. Friction is what makes it difficult. This added friction requires added pressure (boost) that never benefits the power plant. It is consumed in the process of moving the air. Just like we exert an effort to move Coke through a straw, this additional boost consumes energy to produce. The elimination of this friction is what leads to higher efficiency. If technical explanations scare you, skip to “conclusions” at the end. That may be primer enough to return and attempt to understand the explanation.
A SHORT LIST OF TERMS:
-ACHE-Air Cooled Heat Exchanger
-CAC-Charge Air Cooler, aka intercooler. An ACHE that reduces the temperature of the compressor output air product.
-CIT-Compressor Inlet Temperature, air temperature entering the compressor.
-COT-Compressor Outlet Temperature
-CIP-Compressor Absolute Inlet Pressure, normally barometric or slightly less.
-COP-Absolute Outlet Pressure
-CHOKE-compressor operation off the right side of the compressor map where no added airflow is created, only added heat, usually associated by dramatic efficiency losses.
-PR- COP/CIP. As seen in Fig 1, Compressor Pressure Ratio, total absolute outlet pressure divided by the inlet pressure.
-MAP-Manifold Absolute Pressure, pressure sensed at the intake manifold, a MAP sensor, aka boost sensor is used. The measurement is absolute (includes the pressure of the atmosphere). To obtain the boost value, atmospheric pressure must be subtracted.
Note that MAP=COP-DROP
-MIT-Manifold Intake Temperature, the charge temperature exiting the CAC
-DROP-The velocity induced friction loss in boost pressure, between the compressor outlet/discharge, and the MAP sensor, also referred to as pressure loss, pressure DROP or head loss.
-IAT- Inlet Air Temperature (airbox), aka Intake Air Temperature (intake plenum). In an intercooled application we are concerned with airbox IAT. This can be assumed to be the same as CIT.
-WOT- Wide Open Throttle
-PCM/ECM-Power Control Module-the motors brain
-VGT- Variable Geometry Turbine Turbocharger on the LLY, aka variable vane turbocharger.
-Wastegate- work limiting device on some turbos, including the LB7
-Precipitous Overheat- the sudden onset and rapid progression of the overheat, appearing to progress without dampening
-PR-Pressure ratio of the compressor the ratio of absolute outlet pressure to absolute inlet pressure
-PCM-Power Control Module, centralized processor for all combustion related processes.
CONCEPT OF STABILITY
You should understand positive thermal “stability”, and what constitutes a thermally stable platform, and be able to distinguish it from instability.
Stability: a condition where, if an element is displaced from neutral, there exists a force trying to return that element back to its origin. As the element departs from its origin, the restorative driving force against the element should increase in proportion the magnitude of the change. A metal spring is the simplest form to demonstrate stability. The further the spring is displaced, the more force it exerts to return back to its original position, positive stability.
The typical radiator becomes a thermal “spring” after it is warmed up to thermostat temperature. As its coolant gets hotter under load, a larger amount of heat is rejected into the air and the radiator tries harder to return coolant to its regulated t-stat temperature: The growing temperature difference between the coolant and the ambient air flow results in more heat removal, positive thermal stability. As long as the ambient temperature does not change, positive stability will result.
Excessive intake heat is the reason why turbo vehicles lose power
What if, whenever the motor worked hard, a curtain came down over the radiator, in proportion to the motor’s load, gradually blocking radiator cooling air? Coolant would receive less cooling, and temperature, like the bullet, would run away. Thermal in-stability, or negative stability, is a condition created where increased heat production, reduces the cooling ability…this negative thermal stability is a divergent runaway condition. If something interferes with positive thermal stability, it becomes thermally unstable, and this is when “precipitous overheating”, a divergent thermal behavior, can be observed. The following analogy will help.
TAIL WAGGING THE DOG-FEEDBACK EXPLAINED
A funny video clip literally describes a feedback loop; a dog chasing his tail. He sees his tail wag, not realizing that his unconscious commanded a friendly wag. His doggy IQ recognizes the bushy playmate, and begins a progression of trying to catch it, eventually spinning with vigor, faster and faster, until he falls over.
The same thing happens with the heat created in turbo charging. Proposed earlier, what if, when our vehicle got hot, the surrounding ambient air got hotter? Well, nonfictionally, this is what is happening, a curtain of hot air is descending on the radiator, followed by a precipitous divergent behavior overheat, as follows:
The air box gets air from under hood. The workload heats up the motor, thus under hood temps rise, increasing air box IAT. This, in turn, heats up the compressor product, which means that the compressor discharges a hotter air product, with higher exit velocity. The air volume increases with heat up, not unlike a ballon expanding as it warms up. The CAC transfers this added heat to the under hood stream, by way of the radiator, then the fan, to the airbox. IAT goes up further…the cycle continuously repeats. As the cycle develops, the air passing over the radiator is soon too warm to serve the cooling system.
Now that was a mouthful, and it probably sounds like my elevator doesn’t quite make it to the top floor! No worries, this study describes, in detail, the how and the why.
COMBUSTION AIR-FANTASTIC VOYAGE
AIR IS HEAVY! We think of air as nothing, or requiring no effort to move, or compress or whatever. Air is not “nothing”. Anyone who has survived a tornado will know what I am talking about. Air at supersonic speeds will rip the skin right off your body.
An intercooler, or charge air cooler, CAC, acts as an air/air radiator for the intake air, cooling it after the compression of the turbo has caused it to get hot. The compressed, hot air passes through the intercooler, losing its heat to the alloy fins and tubes that form the core. This heat is immediately dumped, with conduction and convection, to the outside air that’s being forced through it by the fan and/or the forward movement of the car.
The induction system has 4 main tasks:
Source: Take cool dense air (ambient air inlet),
Compress: Feed it into an efficient turbocharger for compression and unavoidable heating (red)
Densify: Cool the charge in a suitably matched charge air cooler, (blue) to further densify the charge
Transport: The plumbing between these stages must carry the charge without excessive pressure losses.
1. Source: The air inlet, or intake, should find the coolest (therefore densist) air available; this source should protect the air from temperature rise even when the motor gets hot. The turbochargers predicted performance depends on intake air close to standard day conditions; pressure just above sea level and temperature at 85 degrees, conditions established by the turbocharger manufacturer.
2. Compress: The turbocharger is designed to output a specified volume of air, namely under pressure and adiabatically heated according to immutable thermodynamic laws. This compression heat is a fact of life, but the exact amount of heat is a variable, dependent on compression efficiency. Higher compression efficiency yields air that is heated less.
3. Densify: Now this heat must be removed from the pressurized air charge. This occurs in the CAC, under nearly constant pressure. This cooling makes for more oxygen gas molecules in each fixed volume cylinder “gulp”. Without this step, the turbo serves little advantage to power production.
4. Transport: Plumbing design is an essential part induction. Improperly designed it will add significantly to the overall pressure DROP of the system. As we will see, this additional pressure DROP deteriorates the thermal efficiency of the induction mechanism.
A STEP BACK-PRODUCTION EVOLUTION
Normally a production vehicle is some evolution of a previous model. On the road to keep up with the competition, or to comply with ever-restrictive emission laws, changes are the normal order of things. It can be difficult to see all the possible affects of changes along the path to change, modification sometimes has unforeseen affects, a fitment issue here, an incompatibility there…normally, well thought out changes, implemented with integrity and widespread signoff, go on without a hitch. But no program is perfect, and now and then, a blunder does get by, with consequences that were entirely unforeseen. Consider the possibility that changes made from past developments, could lay dormant as a potential trouble issue. That is until combined with another modification from a recent design change, synergistically brings to life a failure without obvious cause.
ROAD TO LLY
2001- Premiere model Duramax LB7. For 2001, this vehicle was designed to WORK. With most foreseen conditions it did this well, with a cooling systems that protected against 100% of the load that is spec’d by the OEM. It was a little cramped under hood, but that is modern engineering for you. The relatively compact grill exposure did seem a bit small, perhaps more wife friendly than work friendly. It’s only something you might notice if you were a cooling system engineer.
2002- Back then, the airbox design was a cold design, which sourced air from the fender area where it was not overly heated by the motor. There were 2 small problems. The narrow openings in the fender made for a significant restriction, and so a lot of suction existed. Any water or debris making its way to the fender area, found its way into the element during high load conditions. Somebody was traversing a tropical rainstorm, using a dirty air filter element; the element got wet, collapsed and became a major restriction to flow. The resultant codes threw the PCM/ECM into a fit and created an engine “limp home” mode. Eventually this happened in numbers. Second, GM was using a relatively un-reinforced pleated element, and with the restriction that water or debris would create, the tall un-reinforced pleats would collapse on itself, creating more resistance and catastrophic element failure. GM responded with an engineering tech bulletin, and an EC for future production. They sealed off the inner fender openings, opened up the box, and spec’d a new reinforced element for the high flow diesel. Now, sourcing heated air from under the hood, the new air box would do ok, if the fan stayed off. Upon fan engagement, however, the increased pressure of the fan displaces the cool air in front od the airbox, and preheated cooling stack air is forced into the intake. Shown here is 186 degree air going into the air box, at 50 mph, preheated compliments of the fan-coupled hot air design. (I have personally seen 235 F.) Fixing this phenomena would prove daunting.
(These scan gauges give a comprehensive peak into the PCM that other gauges cannot equal, they also read and clear OBD codes)
GM was probably aware of the disadvantage of this change, but it was not any real concern to the engineer who knew full well that the turbo didn’t much care, it (LB7) had a waste gate, after all, hence a limit to performance and performance originated heat. No matter what, it will prevent the production of boost greater than around 20 psi, and limits the negative thermal consequences of such an intake modification. So the positive stability is, in a sense, preserved by the turbochargers safety net, the waste gate. This is explained fully in the section titled “HOTTER CHARGE AND THE ECM’S RESPONSE”, later.
2003- EPA and the Clean Air Act mandates future changes to diesel design in order to meet stringent emissions requirements for 06 and beyond. GM decides to play it safe and be compliant in 04, ahead of schedule. Among several “changes” plans include computerized boost control, exhaust gas recirculation, and catalytic exhaust treatment.
2004.5-LLY debut: The resultant motor replacing the LB7 was the “evolved” LLY. As mentioned, a new turbocharger based on variable geometry turbine stator vanes, VGT, for boost control, and for the first time, PCM controlled boost were introduced. The demise of the waste gate will become very important. Also, it is not known why exactly, but GM chose to reduce the size of the fan pulley, increasing fan speed. Perhaps it was in anticipation of higher heat rejection needs from the above changes. We’ll look at that, but by now, nobody really took notice of the intake air temperature problem introduced in 2001, so the “hot” air box went without change for 04, a critical decision. EGR was also introduced, but the EGR itself was not a big deal. However in doing so, and most importantly, the new intake plumbing had to be sized down to fit into a smaller space to accommodate the EGR, a fatal mistake that left the induction tract chock full of new restriction.
It is this new compressor mouthpiece that is the main folly. This will become critically important, as it will combine with the earlier evolution changes, and will become apparent later why this was, of all things, a cooling system mistake.
LLY Compressor Inlet Mouthpiece
That’s the history, and I recognize that none of it makes any sense yet. Don’t worry, it will, as promised.
TURBO EFFICIENCY-TASK 2
I said earlier that the turbo heats the air, an undesirable side effect of compression. When air is squeezed together, compressed in the turbocharger, work is performed on the air. All the air molecules are forced closer together, and the temperature rises according to laws governing the relationship between Pressure, Volume and Temperature. PV=T. If P is increased then T is increased. The exact amount of heating is variable, and depends on various factors that stand to make compression more difficult. If the induction system is perfect, 100% efficient, then the temperature rise is exactly as predicted in this formula, the minimum amount of heating. But in our compression process, we have less than a perfect process and the T rises more than predicted in this simple formula. The more the air is heated above that predicted by adiabatic compression, the less “efficiency” the compressor has, and vice versa. These efficiency numbers are represented as map “islands” in Fig 1.
Lower efficiency means greater heating. Hotter air takes up more space than cooler air, so this means a higher volume of air must get through the plumbing (task 4) in the same amount of time, it has to move faster, which, in turn, means more friction and resistance.
Geek’s Speak Notes: The analysis for this increased compressor out air velocity and the resultant pressure loss, is rooted in complex adiabatic compressible flow equations, and will not be reproduced here. The simpler concept of the ideal gas law, PV=(nR)T is complicated by the compressible nature of gaseous fluid flow under such dynamic conditions. (nR) is a constant and by and large, the general principle of (added heat=volume expansion) is governing the process. When you heat a parcel of air, it expands when pressure is held constant, and PV=T becomes V=T. When volume is held constant, like in the compressor, P=T. Known as adiabatic heating, when pressure rises, temperature rises. But the process is finite and thermodynamically irreversible, and volume increases also. The greater volume must move faster through the same conduit (CAC plumbing), creating more friction and turbulence, thus also demanding more pressure (force) from the turbo to overcome this added flow resistance. Deterioration from another feedback cycle.
Fig 1. Garrett Compressor Map
Fig 1 shows the compressor map of a similar turbocharger. The Y axis, pressure ratio, PR, is the ratio of outlet and inlet pressures. The X axis represents the mass or weight of air that is pumped through. The oval islands represent areas of equal efficiency. Note the compressor rpm is shown in arcs, and that the max published velocity is 116,923 rpm. Anything above this represents overspeed and unanticipated excess stress to the compressor. Anything to the right of the map represents the choke region where no significant additional airflow is possible, although charge heat increases at a dramatic rate. Longevity expectations take into account maintaining within these limits, and turbo manufacturers require that 3rd parties refrain from applications that do not limit use to within this envelope.
Created in lab conditions, this map is correct for the new charger, in stated ideal conditions, usually specified by the manufacturer in this case, 85 F and 28.4” Hg, which is near sea level. The x-axis is valid only for these conditions. When intake temperature increases, volumetric air flow remains nearly constant, but the corrected mass air flow xaxis shifts to the right. More simply, warmer air has less weight for every compressor “gulp”. So for a given PR and motor rpm, volumetric flow stays constant, but mass air flow decreases as inlet temperature increases. If this isn’t clear, don’t worry, it’s not too important to understand. It is important to know that these departures from “perfect” conditions have the effect of reducing compressor efficiency, reducing oxygen density, the opposite of what the turbo is there for, and the statement
“Excessive intake heat is the reason why turbo cars lose power”.
should have more meaning now. Excessive intake heat means less useable oxygen. As I’ll show later it also means the turbo is harder to drive, so more engine HP goes to spinning it. Less goes to the wheels. When the turbo is harder to spin, the exhaust backpressure, or turbine drive pressure, goes up and EGT increases.
Author’s Technical note: In the examples that follow, I have decided to illustrate each condition on the same map, overlaid for comparison purposes. This requires the reader to understand that the x-axis, corrected mass air flow, is not valid (except for the rarely seen ideal 85* F IAT), and shifts to the right in each case, as per the increased IAT. It would be more correct to label it volumetric flow rate when dealing with changing inlet temperature conditions. I could have left this out, but someone would have picked it up.
I spoke about the heat that goes into the charge air during compression. This happens because the air is being “worked”, not unlike rubbing your hands together creates warmth. That means that excess heat is created in correlation with the imperfect nature of the process, the less perfect (less efficient) the more heat is produced. Any heat that is produced reduces the density of the air, so we need to minimize this heat production during compression. Exactly how much hotter the air gets as it is being compressed depends mainly on how much it is being compressed (pressure ratio), and the compressor’s (adiabatic) efficiency, η. The emerging compressor outlet temperature, T2 or COT, can be calculated from 3 changing variables: the compressor inlet temperature, T1, the pressure ratio across the compressor, PR, and compressor efficiency, η.
T2=T1+ heatup, or
T2=T1+ (T1 x PR0.286 ) / η
We want the lowest possible compressor exit temperature, T2. This will happen by minimizing the right side. So we need the lowest possible inlet temperature, T1, AND the lowest PR, AND the highest compressor efficiency, η.
For clarification, 100% efficiency does not mean that heat is not produced. It only means that the minimum possible heat is produced, an important distinction. 80% is the practical limit in our application; seldom do efficiencies exceed 74% in max power. In extreme cases of over boost, our turbo is occasionally observed under 60%, with 68% (.68) being common. The extra heat of this condition is very significant, as in the following example.
Assume IAT, T1, to the turbo is 100 F (560 Rankin) and the boost pressure is 28 psi. Pressure Ratio = (28+14)/14=3.0) the theoretical outlet temperature, T2, will be:
T2=T1 + ((T1 x PR0.286 )-T1)/ η
T2 =T1 + (560 x (3.0)0.286 -560)/ η
T2 = T1 + 207 / η
This means that there is a temperature rise of 207 F, in a perfect process. So exit air is 307 F. But now factor in efficiency less than 1.0 (100%).
If we assume a typical compressor efficiency of 68% (.68):
207/.68=304 F rise, or air exiting the compressor at 404! Whoa, a 48% increase in heat due to efficiency alone! It is important to control IAT AND efficiency!
HOTTER CHARGE AND THE TURBO RESPONSE
OK, we got this far, the meat and potatoes. Reviewing, increasing IAT causes the temperature sensitive x-axis to shift to the right, so that for a given operating point, mass low of air decreases (volumetric flow remains constant), a bad thing for power. Also compressor discharge temperatures increase by 140-175% of the IAT increase, which creates yet more heat in the CAC, a bad thing. For constant boost conditions, a 100* F degree increase in IAT (CIT), increases discharge COT 140-175* F degrees. The added heat expands the air flow volume, basically forcing a higher air velocity in the volute and in all the intake plumbing after the compressor, a bad thing. This speed increase creates additional plumbing pressure DROP, friction, which the LLY ECM will now dutifully compensate for by commanding an increase in compressor outlet boost, a very bad thing that will lead to progressive cyclical heat up described earlier. The LB7 could not do this because boost was not computer controlled, but rather waste gate limited. The LLY has closed loop boost control with the MAP (boost) sensor acting to regulate intake plenum pressure. In the earlier LB7, the same MAP sensor had only a reporting function and did not schedule boost, fuel or any other engine functions. Sometimes it is best to not fix something that is not broken.
Here is how this divergence takes shape. The increased charge air temperature is getting rejected at the CAC, and since it is getting hotter, so is the air in front of the radiator, coming off the CAC. The thermo-viscous fan comes on, reacting to this greater temperature. With it, the IAT climbs because the fan directs this hot air right into the air box, where the cycle repeats (dog spinning faster) and creating negative stability to the cooling system. As heat soak overwhelms the CAC, a curtain of hotter and hotter air emerges in front of the radiator. The radiator cannot cool as well…the fan is on more…the IAT goes further north…leading to greater CAC heat soak…ad nauseum: the birth of thermal divergence…negative thermal stability.
Now with the dizzy dog the fix is easy, just put a blindfold on him, cycle broken. The feedback is removed. Similarly with the induction system, there are 2 things that must be fixed simultaneously. 1. Remove excess sources of restriction in the induction tract, and 2. isolate the air intake from under hood temperature influence. Sounds easy? Now before you go off saying “I have a CAI and it doesn’t fix my problem”, let me tell you that most advertised CAI effectiveness is about as short lived as a snow flake here in Phoenix. I have seen “cold air” intakes produce IAT as high as 190 degrees. They range anywhere from pathetic, to unworthy of the landfill. There is no bigger fraud in the aftermarket industry. To truly isolate ambient from under hood, you must go outside the engine compartment for air, and insulate the box and intake plumbing. But let’s not digress too much. It is remotely possible to overheat the vehicle even with a good CAI; it is just much easier and quicker to do it with a hot air intake, some inefficient turbo charging, and poorly designed plumbing.
BY THE NUMBERS, FROM LB7 to LLY
95 degree day, truck warmed up, no load, under hood temps a modest 25 degrees over our ambient 95 degree day. The LB7 warm design airbox is now taking in 120 degree air. Then we put the pedal down to the wood, and introduce a load, say pulling a large trailer into a headwind. The turbo spins up, and pulls 120 degree air into the airbox.
Fig 2-LB7 WOT (GREEN)
The LB7 creates 20 psi at the plenum, with 21 psi at the compressor discharge. Normally this would mean a compressor output temperature of about 340 degrees. The 1 psi plumbing losses require compressor discharge product at 21 psi, the PR is (21+14)/14=2.4, and we are at 73% efficiency on the map of our used turbo. Typical… and routine.
Fig 3-LLY WOT (yellow)
Needto correct for inlet restriction
Now for the LLY. The LLY PCM is requesting 20 psig via computer signal, and the turbine vanes get adjusted to make it good. The compressor discharge is 23 psig due to the increased hardware plumbing restriction (smaller tubes to/from the CAC), compared to the LB7. Additionally, the intake resistance prior to the compressor has increased because of the new smaller intake confinement. There is a 4 psi loss just prior to the compressor. This change to the vehicle is so detrimental to turbo performance, that I have dedicated an entire article to it at THIS LINK.
The PR is 3.6. (correct chart for new PR) Remember the formula:
T2=T1 x PR0.286
T2 goes way up with the higher PR. Is plumbing resistance important?
Fig 4. LLY Example of stock 20 psi intake boost (23 psi at the compressor discharge), with 160 IAT, 40 mph vehicle speed. Compressor discharge temperatures up to 450 degrees COT, radiator ambient cooling air up to 170 F. Note the gradual feedback climb from 140 to 170. Compressor Efficiency calculated, 68%
This is not all geek theory and postulation nausea. Fig 4 is actual test data of a stock LLY vehicle with a compressor discharge boost of 23 psi (20-21 psi at the plenum), which includes 3+ psi of DROP. The red/yellow is a redundant reading of compressor outlet temperature, COT. At the beginning of the load, COT is about 400 degrees, COP is 23 psi. This steadily rises to 450 F and 24 psi, as the feedback develops. See the gradual climb? If that 3 psi of pressure DROP was reduced to 1 psi, by virtue of a free flowing design, the same plenum boost could be obtained at 21 psi compressor discharge. IAT would be less, as the feedback loop strength is reduced, the compressor would be operating at 69% instead of 68%, and the COT would be about 50 degrees less. Less heat, reduced turbo rpm, longer life, just by increasing the plumbing size (options discussed later). Also, see how ambient slowly rises from 150 to about 170 F? That is the temperature of the air when it arrives in front of the radiator; the result of increased heat rejection by the CAC, partly due to heat soak, but mainly from the divergent feedback cycle. Toward the end of the 1 minute run, the radiator is being cooled by 170 degree air.
23 PSI FOR 20 PSI, WHERE DOES THAT 3 PSI G0?
Into overcoming friction after the compressor. And if you are still thinking that “air doesn’t cause friction”, consider then how the titanium skin of the SR-71 climbs to 220 degrees when flying 1500 mph. That friction creates very large pressure losses. Air is flying through our CAC plumbing VERY FAST, over 200 mph, and this comes with a price: friction and pressure DROP as mentioned earlier. The faster the charge must go, the higher the frictional pressure loss. In our case that is 3 psi that the turbo must produce, that does not contribute to plenum boost! Worse, it is converted to heat and can be measured. A turbo can provide 40 psi of boost, but if 10 psi goes into plumbing restrictions, you still have only 30 psi to the cylinder, but it is as hot as the inefficient 40 psi product. Is plumbing important? It’s huge!
Representing the square relationship of pressure DROP to velocity, typically if you double velocity, you 4-fold the pressure loss. This is no different than wind resistance on a vehicle. Well, as it happens, one other thing GM did in welcoming the new turbo, was shrink the diameter of the CAC tubes from 3” to 2.5” OD! It could be they wanted a lower volume system to spool up (pressurize) faster, or possibly it was a botched backpressure design for the new CARB standard, or a mere $0.75 cost saving. Regardless, this is a mistake for us. The CAC itself is the same, with 3” openings. So I am left with more questions than answers on this one, but there is no question, it was a bad choice.
If you do the easy math, that change gave up 35% of the flow area, and forced the charge to go 53% faster (using 2.1” and 2.6” ID respectively). Consider the LB7 at a stock airflow rate of 700 SCFM, its compressor discharge tube (CAC tube) has a tube velocity of 160 mph. The LLY is in excess of 240 mph with its smaller plumbing! This increases the pressure DROP in the tubes, by over 1.5 psi total. The LB7 system had a TOTAL plumbing loss of 1-1.5 psi including the CAC. Now the LLY has over 3 psi loss, due to this plumbing change, generally an unacceptable compromise for faster spool. So to generate the 20 psi that the PCM calls for at the intake plenum (boost sensor location), the LLY turbo must perform 23 psi at the compressor outlet, the cost of under sizing the plumbing, as mentioned earlier, more speed induced friction. This number has been verified, and is consistent with adiabatic compressible flow calculations. This is under “good” conditions now, 120*F IAT.
Here is a closer progressive look at this cycle.
CHAPTER 1: BOTTOM OF THE HILL
In Fig 3, at 23 psi, the PR on the chart is 2.7, and red line rpm of 3200, 43 lbs/min (corrected), the efficiency of the 120 IAT (cool) turbo is now 71% and the output product has climbed to 380 F, a 20 degree increase. While not a staggering change, this hot product now goes to the CAC to get cooled. The earlier 360 degree product would have been cooled to 170. The pressure DROP laden 380 degree product gets cooled only to 175 F. Result: A bit less O2 and power, higher EGT, and more heat being rejected by the CAC, into the radiator. By itself, this is not a tragedy, but this is just chapter 1, early in the cycle. ECT is just starting to rise, and now the fan comes on.
CHAPTER 2: NOW THE FAN
Consistent with modern compact packaging, the fan is literally 1” from moving belts and pulleys, with a cast iron motor right in front of it, reducing airflow and impeding flow out from under hood. As a result, much of the superheated cooling air is forced up under the hood and the air box can’t choose where its supply comes from. The smaller fan pulley for the LLY means a faster spinning fan creating more parasitic drag. Since the fan drive ratio was already optimized on the LB7 for this fan, the smaller LLY pulley did 3 things:
1. Made it louder,
2. More of a parasitic power loss (drag),
And most importantly,
3. Better air box heating mechanism, explained here:
The fan is designed, such that, when rotating CCW, the hottest air coming off the cooling stack, is being swept upward by the blades on the driver side (hot side), and rotationally dumping right onto the airbox on the passenger side. You can feel this even with the hood up. This is attributed to the cramped quarters of the original 01 design. The new fan pulley just does it louder, and with more force! Since the airbox is fed pre-heated air, we now have a completed feedback mechanism for negative thermal stability. So much so that when the fan engages on a grade at WOT, IAT has been observed to rise from 120 to 200 in less than a minute! So let’s go through the iteration again now with a conservative 180 degree intake air. With higher IAT, the compressor discharge is hotter still, expanding the pressurized product, increasing velocity to 340 mph, so plumbing pressure losses increase another 2 psi. Now the compressor must deliver 25 psi at the discharge to compensate for the 4 psi loss, the output product is 500 degrees and efficiency is reduced to 69%, the cost of higher boost. In Fig 5, the compressor now approaches overspeed rpm conditions, and off-map operation.
Fig 5-LLY after extended WOT operation with fan (red)
Thermal feedback resulting in reduced efficiency/increased rpm
Fig 6-altitude effects (magenta)
CHAPTER 3 ELEVATION EFFECTS
By now we have climbed to 5000 ft (12 psi baro) of altitude, and the problem compounds further.
PR= (26+12)/12=3.17. The compressor is now running outside of its published allowable envelope, and a tuner has not even been added yet. Did GM plan this? Garrett was probably never consulted about an operation off the approved envelope, effectively over speeding the compressor in every tow trip where IAT increase is experienced (essentially all summer trips).
CHAPTER 4: THE LAST STRAW-BOOST TUNING
This photo is what an over boosted turbo looks like after a 10 minute cycle of full throttle. A glowing red compressor is indication that the cycle has a firm foothold, it adds lots of added heat for the CAC to deal with. Take a reasonable 85 HP tune which adds 2 psi boost, add an additional 5 psi performance boost enhancer onto it. Run it up a hill on a hot day, load in trail, such that WOT is required for 5 minutes or more, at an elevation of 5000 ft (12 psi barometric pressure). In this case the WOT boost at the intake plenum is commanded to about 29 psi. This amount of air flow produces 5 psi of plumbing pressure losses. This means the compressor must manufacture 33-34 psi. After a minute or so at WOT, the premium air flow rates seen with this combination, in conjunction with the thermal rise seen in IAT, the plumbing pressure loss exceeds 5 psi, then 6 psi. Now the compressor discharge is manufacturing 35-36 pounds of boost. Choke levels! Now look at Pressure Ratio on the map. The higher altitude brings y-axis PR up to, (35+12)/12=3.9. See Fig 7.
Fig 7: 7 psi of Boost Augmentation/tuning plus elevation (black)
This is clear operation in the choke region. This means that there is no more actual airflow, just added heat. With a resulting IAT of over 220, the compressor is in the 48-52% range of efficiency, and the compressor discharge is 675 degrees!!! Nasty and insane, don’t expect the turbocharger to last much longer.
WARNING: PROFESSIONAL TEST RESULTS, DO NOT TRY THIS AT HOME!
An attempt at humor, realize that this extreme test represented by fig 8 is a very dangerous event to the integrity of the motor. An innovative tuning program, EFILive, allows me to safely scan all system parameters without unintentional disasters, then tune for safer operation. It is possible for the operator to combine equipment (stack) in an unintentional meltdown. This is represented by the black line of Fig 7. It demonstrates the tremendous effectiveness of the CAC, but it is also testimony to its ability to erode radiator cooling performance. The CAC is a heat exchanger, ACHE, much like the radiator. Each needs an abundance of air, cooler than the medium it is trying to cool. The by-product of the CAC is ambient heat. The cooling air is dramatically warmed up by the CAC. Yet, this same cooling air is expected to satisfy the needs of the radiator after it leaves the CAC. When it is this hot, it will not cool anything.
For this test, the CAC heat rejection calculation using the resultant 55 lb/minute of charge airflow, 0.25 BTU/lb-F heat capacity, Cp, and the resulting 400 F CAC temperature reduction (590 to 200), gives a CAC rejection of over 330,000 BTU/hr! This is enough heat transfer to heat ten 5- bedroom homes on a 20 degree day. The radiator, sitting behind the CAC, receives air in excess of 220 degrees… FOR COOLING, as shown in actual test data in Fig 8. It can not possibly function to keep coolant contained below this. In reality, the CAC is actually heating the coolant and precipitous overheat follows. In about 60 seconds, ECT (not shown) rises from 190 to 260 during this verification test.
Fig 8. Compressor chokes. 31 psi of intake boost (36 psi at the discharge) creating a precipitous overheat condition in one minute. Note what happens to COT when the fan comes on (white line). Note the ambient temperature (blue) in front of the radiator tops 250 degrees at 40 mph. Compressor PR 3.57, COT 590 F, compressor efficiency 61% (must be calculated in this case)
Of interest, look at what happens halfway through the WOT run, represented by the vertical white line. The COT increases suddenly by 40 degrees and the ambient decreases commensurately. What has happened at this point? Speed, RPM, throttle position, and grade all remain constant. The fan came on, increasing IAT which, as we know by now, increases COT even more. This is a nice clear pictorial of the negative feedback mechanism at work, as fan induced IAT (not shown) rises from 130 to 170 in a few seconds. So the fan actually created a hotter CAC! In certain cases, the fan creates more heat than it removes. After the fan creates more airflow, the ambient is still blazing at a slightly reduced 230 F. No radiator has a chance being cooled by 230 degree air, and this truck overheated in less time than it took to notice it was getting hot. Fig 8 shows the impact of large pressure DROP, at 5-6 psi. This can be reduced to 2.5 psi with plumbing improvements. Doing so would drop PR to 3.4, reduce COT from 590 to 500, and increase efficiency to 66%. The turbo will spin 15% slower, and that will greatly extend its life (though still operating out of the envelope). But most importantly, the vehicle will have more HP with the cooler charge, more forward speed for cooling, and this OH scenario is improved.
So what does 220 degree intake air do? It kills 20-40% of your power for one (see timing discussion later), and also cooling capacity. That is, the ambient cooling air coming through the grill rises to over 200 degrees (250 degrees has been logged in speed limited trials) in front of the radiator. The radiator cannot do its job, so ECT rises steadily, keeping pace with this deteriorating thermal situation. Note also that a new radiator with more “capacity” will have exactly the same problem, this is not a radiator design issue, and cannot be fixed with any type of radiator redesign, or a piggy back radiator.
Also the combustion air that leaves the CAC is also very hot, up to 200 degrees over ambient. 300 degree air has less oxygen, and burns faster, a timing consideration.
THREE PILLARS OF INDUCTION OVERHEATING
By way of review, Thermal Feedback requires 3 elements:
1. Overuse of boost in a poorly designed, undersized induction network
2. A negative stability mechanism through intake air preheating, and
3. Limited natural cooling airflow, to stem the effects of 1) and 2), an inherent flaw in aerodynamic design considerations dating back to before 2001.
With some familiarity now, use THIS CALCULATOR and play with the numbers. Pay particular note to part 2, where the calculator predicts how much radiator function is lost due to looped IAT heat up.
I WANT MY WASTE GATE BACK!
The main function of the waste gate in the 01-03 LB7’s IHI turbocharger, is to put a cap on what actual boost the turbo will produce. I can’t overemphasize how important this is. It limits how far into the turbo map this thermal feedback loop can progress, regardless of plumbing losses. It is the proverbial “blindfold” on the dizzy dog. There is no such protection with the LLY, although PCM software has this potential as a software waste gate equivalent.
LB7: Mechanical waste gate limit of 19-20 psi. That is all the work that the turbo will do, regardless of how warm IAT gets. Simple work limit design that regulates the pressure at the compressor exit (COP). Simple, reliable, smart…effective.
LLY: Computerized set point of 19-20 psi. no waste gate. By contrast, this computer controlled turbo charger seeks to do one thing, generate boost as needed to satisfy the PCM’s request for a specified MAP pressure AT THE INTAKE PLENUM! If there is 10 psi of plumbing resistance loss, requiring the turbo compressor produce 40 psi at compressor outlet, to make good the 20 psi at the intake, it will perform this task like a soldier. And that is the problem: there no cap on turbo efficiency degradation, no cap on the heat produced by the turbo, thus no “control” to prevent negative thermal stability. Restrictive plumbing adds an element of steroids to the deteriorating condition, by way of reduced diameter intercooler tubes, and poorly designed intake plumbing, discussed earlier. And if this is not enough bad news, there is actually a provision in the OEM programming to increase boost as IAT rises: imagine steak sauce on the dog’s tail!!!
So what CAN be done?
Given the design limitations, boost reduction at the higher RPM’s is essential. In my opinion only 1-2 psi reduction at the highest RPM settings in the PCM’s boost correction table when these thermal conditions are encountered. Enter the powerful tuning tool, EFILive, correcting boost back in level, only when these conditions are encountered. Combined with larger CAC tubes and a trusted cold air intake, this cycle can not exist, and precipitous overheating ends.
JUST ONE MORE THING-TIMING
Consider injection timing. It has not been discussed much in the effort to keep the discussion simplified: the affect of increasing manifold Intake Air Temperature on optimum timing. In studies of DI diesels that vary IAT, one shows the peak heat release rates being accelerated 20 degrees with an IAT increase of only 70 degrees C. And this was at an RPM of only 800. This means that increased charge heat dramatically reduces ignition delay, and speeds up the diffusion burn inside the cylinder. Hence, in the absence of a correlated timing correction (retard), much more of the burn occurs prior to TDC, reducing motor efficiency in a huge way. The result is usually noted in the form of reports that the customers mpg goes way down, with a concurrent loss of power. Does this sound familiar? If you have ever overheated, and were paying attention, there is the gradual feeling that you are losing power. This is the reason. Poor oxygen density and over-advance timing.
Under the existing stock timing maps, we see as much as 13 degrees BTDC during loaded operations, as much as 20 degrees with some aftermarket tuners. Along with more negative torque before TDC, this also sharply increases internal cylinder pressure to harmful levels, a longevity concern. Hence to preserve power in the torque stroke, timing must be retarded under high load conditions, as intake temperatures increases. Yet there is no provision for this. Timing will hang out severely advanced right up until coolant is on the ground, and IAT’s topping 230F. A defuel schedule does attempt to mitigate this, with ECT’s beginning at 244 F, but does little about the negative feedback induction loop that creates this high IAT. At 244 ECT, this loop has already progressed deeply into an overheated CAC, with IAT’s often exceeding 200. Unless you crest the hill by this point, nothing can be done, except to pull over.
Fortunately, provisions to prevent the runaway condition on the ECM/PCM do exist. I have had some success with it, and continue to familiarize myself with the concepts, and anticipate an ECM release that will allow the truck to be more fully utilized with this type of compensation, a “hillbuster” option, that will not permit overheat, with a gain in power and mpg as side benefits. The main limitation with this method is the increase in EGT that comes from reducing boost.
Using actual hillside load test results, combined with predictive modeling, improvement is demonstrated. The stock tuning is used in a full throttle climb test on an 8% grade, using the stock warm air intake. IAT is seen to rise from 90 degrees.
Stock average climb speed….48 mph, peak IAT 178
With timing compensation….51 mph, peak IAT 155
This is with no hardware changes. This 3 mph difference may not sound like much, but it represents a 20% power increase. The added frontal pressure from this forward speed difference means better cooling.
The 4th Pillar of induction overheating:
4. Heat induced ignition advance, causing loss of power and motor efficiency.
Incidentally, EGT monitoring will fool you into believing that you are running cooler. If you find yourself disillusioned that EGT is the main indicator of future ECT, retrain your thinking, it’s not. It is incorrect to believe that because EGT is down with excess boost, that you have improved conditions for the cooling system. The exact opposite is true. Reduced EGT through augmented boost means only one thing, excess ambient air dilution in the combustion chamber. That cooler (and uncombusted) air is cooled in the CAC, that extra heat spilling on to the radiator. If you want reduced EGT, seek it elsewhere, this method creates problems, not remedies. EGT is a sensible parameter to monitor, in fact EGT is a thermodynamic necessity for power production and should be monitored to protect hardware, but not as a cooling system gauge. I highly discourage the purchase of such devices whose sellers claim “reduced EGT”. You will be doing so at the expense of turbocharger longevity and cooling system capacity. There is no way around this, as long as the CAC sits in front of the radiator. If that is not compelling enough to reshape thinking, then consider this paradox: They want you to believe that reducing EGT via excess boost is good for the turbo, yet never mention that the added boost amounts to off-map overspeed rpm and unapproved mechanical stress. Yeah, right!
A DIFFERENT ANGLE
How about some other thermal realities about the air that cools behind the grill? At sea level, the 4.5 square foot cooling stack can muster about 6000 cfm of cooling air at normal speeds. If ambient air is 100 degrees, and it can rise to 220 degrees, as defined by maximum acceptable oolant temperature, then the stack has a capacity of about 750,000 BTU/hr. At altitude where air is thinner, mass air flow through the stack is less, about 20% less at 5000 ft. With the temperature decrease, the net result is about 5-10% less cooling capability, or 675,000 BTU/hr tops. The thinner air also aggravates the thermal feedback loop, because PR (see map) increases with altitude, as boost is held constant (which the LLY does). At sea level, with contained IAT, the CAC will reject 120,000 BTU/hr at full load, consuming 20% of stack capacity (see fig 10). At 5,000 ft, after IAT has run away to 220, plumbing velocity is 390 mph, pressure DROP totals over 5 psi, the CAC is rejecting over 350,000 BTU/hr, consuming half of stack cooling capacity now. Use this RADIATOR CAPACITY TOOL and see that for all practical purposes, the CAC becomes an ambient air heater to the radiator, gradually negating its capacity. By way of comparison, the tranny and AC condenser contribute 20,000 and 30,000 BTU/hr respectively. The remainder is all that is left for the radiator. Not good, since the radiator must exchange nearly 450,000 BTU/hr to safely serve the combustion process at 100% duty cycle.
If you are still with me, then maybe you won’t mind a couple more charts. The following show the overheating LLY, and then show the (dashed) modified LLY which has all the induction overheating pillars removed. Specific changes are a “restriction-free” cold air intake which utilizes an external fender scoop for ambient pickup. Also changed are upsized boost tubes, and the stack sealing/shrouding kit. In addition a revised PCM tune is used that effectively creates a software “waste gate”, inhibiting the looped CAC heat up process. No cooling system expansions are installed.
Fig 9. LLY Induction Overheat
Look carefully at what happens to the stock truck when the fan comes on after 16:15. IAT shoots skyward. ECT appears to start a decline and then the increased induction heat exceeds the fan benefit, and ECT begins to rise again. Then note the steady IAT rise, dotted, from 175, steadily rising to 225…thermal feedback: a continuous loop of deteriorating cooling performance, and power reducer. Look now at the dashed performance, with the improvements. Complete containment, higher speed, taller gear is maintained, specific fuel consumption has decreased, the turbo is spinning 25,000 rpm slower, the charge air is delivered to the motor 85-90 degrees cooler, EGT is nearly 150 F less. This is a truly remarkable transformation.
Fig 10. CAC-Induction heat rejection
This chart show the quantitative impact of the thermal feedback, the calculated heat rejection from the CAC. The heat produced in this example is more than cut in half, from 250,000 BTU/hr to 100,000. Many things can be concluded from these charts, but a big one is the importance of not running redline rpms. When the vehicle is performing better, and does not downshift, the heat created is minimized.
Look also at what happens to IAT2, the actual combustion charge air temperature: 160 F vs. 250 F. That’s strictly oxygen density loss.
MY INTERCOOLER BOOT KEEPS POPPING OFF WHEN I TOW
Hopefully by now, you have figured out why. The extraordinary amount of pressure drop in the LLY charge air plumbing means that there is around 6 psi more pressure at the compressor discharge than at the intake plenum. This could be as high as a 8 psi difference as IAT heats up. All of a sudden the connection which served 22 psi fine, sees 28 psi and can’t hold on. Incidentally, if this happens, it will always be the drivers side boot, the hot side. It is exposed to the most pressure
SHOULD I REPLACE MY INTERCOOLER?
No! The stock Behr CAC is just fine, a well constructed, premium piece with under 1.0 psi of charge DROP at maximum stock flow rates. Spending $1300 for a .5 psi reduction …. A waste of money in my opinion. The new one will still be subject to all the same problems affecting the old one, until you fix cause. If you live in LA, you can clean the windows as much as you want, but it will never make the smog go away. No, don’t replace the CAC. Spend 80% less and purchase a true Cold Air Intake, and an alternative mouthpiece instead. But be careful, just because it claims cold air performance, does not mean it is true. Most of them are ineffective.
In searching for cause, the focus had been so much on the cooling system, that we didn’t stand back to look at what could be interfering with it. The cooling system is fine, has always been fine. In fact, it never changed in the evolution of this vehicle, save for the overdriven fan.
What is the primary cause of LLY overheating under heavy loads? The redesigned induction mechanism. Once broken down into its small steps, the power eroding process becomes clear. Start with a platform that is already marginal for cooling air passage, take away the efficiency protection of the wastegate, increase the plumbing system resistance to flow, and then create a fan coupled IAT increase. That is the way to insure a non-cooling system related overheat.
This defective process takes engine horsepower and converts it to heat using the turbo. Add some surplus boost to drop turbo efficiency even further, and you have an unstable thermal platform, that can no longer cool itself: a remarkably effective negative feedback loop. With the faster spinning fan accelerating this looped process, quickly the CAC becomes so hot that radiator cooling is no longer possible with the existing cooling air flow. The hotter air charge temps reduce oxygen content and creates a condition of advanced combustion and heat release, reducing power and efficiency further. The resultant power loss and speed decay reduces cooling. Soon, and in cyclical fashion, the vehicle is capacitated. At this late point, if the load is not promptly removed, overheat ensues.
In contrast, the LB7 suffers less from these issues through the use of a waste gate and larger efficient plumbing, which limits compressor output. While the same MAP sensor is employed in the LB7, its use is restricted to diagnostic, not boost control. The fan coupled IAT decay is less also, by virtue of a slower fan that is less coupled to IAT increase. Together, the changes made to the vehicle, in separate evolutionary stages, finally came together in a several development oversights over some years. Fortunately, there are remedies.
To address this, all the elements that support thermal feedback and power decay must be removed. They can not be completely removed, so we must minimize them.
1. Avoid using any tuner that increases boost above stock for extended WOT conditions. I suggest tunes specific to this application, which among other things, should front load fuel and boost together at lower rpm, and automatically adjust boost and timing as thermally degrading conditions are encountered. I have developed safe tunes, using EFILive that add 80-200 ft-lbs of torque, and 40-100 HP which will eliminate the precipitous overheat.
2. If you must tamper with stock boost settings, do so with a variable controller of some type that allows you to remove the boost enhancement for extended WOT conditions. This certainly does not apply to individuals who endeavor to create a 12 second race truck, as this is not enough time to develop these conditions.
3. Obtain an effective CAI, that is free of excess restriction. “Effective” is the important term. Get one that keeps IAT down under the maximum extended workload that you foresee. It must also have a free flowing compressor mouthpiece section. Suggestion here: the big names do not perform this task, not one of them. Most are no better than the stocker. I have supplied one now for 2 years that is factory designed and modified. On first appearance, it looks completely stock, as intended, but nothing else compares to its ability to stem IAT heat up. Compared to names like AFE and KN, it is much more effective, and the superior stock element filtration is used. A fender scoop, like the stinger fender scoop we have approved, is an improvement to the remainder of the IAT heat up issue, but it is not essential.
Note: Alternatively, GM has implemented the 06 CAI retrofit for earlier vehicle. Keep in mind that it only partly resolves the feedback condition. Users typically report only modest benefit. IAT typically exceeds 140 F in these cases, and now you know the significance of this observation. Ironically, GM does not replace the whole intake, and leaves the most damaging piece in place, the compressor mouthpiece. If you really want to see benefits, get it replaced with a free flowing mouthpiece.
4. Upsize the CAC tubes. Like the CAI, this will enhance the efficiency of the induction plumbing system, and lower the heat production from the compressor.
5. Install a nose shroud and stack sealing kit. All of these improvements can be found here on the site.
6. Of course, cooling system expansion is certainly always a good thing, can never have enough thermostatic cooling, and TD-EOC OIL COOLING is the way to go, the beauty of its design resting on the observation that oil temperature leads that of the coolant by 20-110 degrees, varying with sustained load. It is the hottest cooling medium in the vehicle. The TD-EOC, combined with the optional air dam, it continues to reward its customers. It does not fix the induction loop design flaws; it allows the operator to forget that they exist. The true CAI is essential for reducing thermal loop related heating, and is a requirement for power preservation.
Perhaps the most valuable suggestion, it costs nothing. Simply slow down into the sweet spot rpm, 2200. Stay away from extended high rpm, WOT conditions, prone to making the CAC too hot. But if you are like me, you don’t want to be passed by loaded semi trucks, see items 1-6.