IZZITS BATTLING THE. Making the right COOLING SYSTEM

Transcription

1 BATTLING THE COOLING SYSTEM IZZITS BY PAUL WEISSLER Is it this or is it that? The underlying cause of a cooling system problem may not be immediately apparent. Skill is required to sort through all of the possible Izzits. Making the right choices is what sets a good technician apart picking correctly from a pair or more of Izzits ( Is it this or is it that? ). Most Izzits are diagnostic that is, a vehicle may have a problem with two or more likely causes, and determining which one is responsible is not an obvious procedure. Others are more subtle, such as picking the correct part. In midsummer, the focus is on the cooling system, so here are some of the questions you re currently facing. Although no one s perfect, here s how you can make the right choices most of the time. Is It the Belt or the Pulley? We ve heard of shops that replaced belt after belt, then put on a new tensioner, then changed a pulley to try to cure a problem all without success. For so seemingly simple a system, it s amazing how many things can go wrong with an accessory belt drive. If a belt has survived a few years and has stretched so the tensioner is at its limit, or shows fine cracking and small chunks of missing rib, that s normal. If it s oil-soaked, that s another issue, with another root cause and another solution. However, a belt that shreds or wears prematurely or jumps off the pulleys either is the wrong belt or is likely to be suffering from pulley misalignment. And in today s tight under- hood environments, misalignment is tough to see unless the pulleys wobble more than a belly dancer. Just because you can t see a pulley wobble, that doesn t mean it s properly aligned. The pulley could be running on a slightly bent shaft, or the accessory mounting or its bracket could cause a tilt. There could be horizontal misalignment, as well that is, one accessory slightly forward of another. In any of these cases, there might not be any looseness in the pulley you could actually feel. A new alignment checking tool that uses laser sighting is being marketed by the Gates Corp. to provide the diagnostic assurance technicians need. The tool is designed for ribbed pulleys. Af- 24 August 2003

2 Photos: Paul Weissler Turn over a ribbed belt at many points along its circumference to see if there s any damage other than hairline cracks or a few small chunks missing. ter making sure the laser sight is properly aligned (with a cardboard strip gauge provided), remove the belt from the engine. Then install the tool in one of the pulleys, turn on the laser and rotate the pulley so the beam hits other pulleys (safety glasses are provided in the kit). The beam should hit the midpoint of any other pulley, but there s an acceptable tolerance, depending on the distance between pulleys within a half-rib for a six-inch spread, a full rib for a one-foot spread and up to two ribs for a two-foot spread. In addition, of course, you should inspect pulley grooves for debris, trying to rock pulleys to check for looseness, turning them to feel bearings for roughness and checking the spring tensioner with a wrench to feel for any weakness or lost motion. Also inspect the flat side of the belt for glazing or damage, since the water pump usually is driven by this side. Wrong belt is a tough call. Although we occasionally see a misrouted belt, or a belt installed one rib off, more likely a noisy belt or one that pops off (assuming the pulleys pass inspection) is wrong for the application. In some cases the belt is the right overall length, but has one rib fewer than it should. That s the wrong belt, plain and simple, and even if the perimeter fit of the belt is good, the belt is marginal for power transfer prone to slippage and squeal. A more subtle problem is a marginal fit. The marginal size belt may be an acceptable fit on most engines, but not on all. This isn t to say that the OE belt is always the right choice, but if a replacement belt gives a problem, going to the original equipment source may fix the problem. If you ve verified that the pulleys and tensioner are aligned and in good condition, you re making a part replacement in which you can have reasonable confidence. Is It the Radiator Or the Water Pump Or the Thermostat? Here s the scenario: The engine is running very hot, maybe overheating, in slow-moving traffic, but cooler in highway operation. There s no debris buildup on the condenser, but the August

3 BATTLING THE COOLING SYSTEM IZZITS cooling system itself has been neglected. Is the radiator restricted or is the water pump weak? You used to be able to look into the radiator through the fill cap and perhaps see a buildup of corrosion on the tubes, or debris in the tops of the tubes. And you could lower the coolant level in the radiator and run the engine, looking for adequate coolant flow from the engine, which is a quick check of the pump operation. With today s engines, the radiator is not the high point of the cooling system, so the fill cap is elsewhere and there s usually no practical way to look inside the radiator. You don t want to make a mistake with this problem, because water pumps often are recessed into the engine. Even if they re not, they re anything but a quick replacement. However, even though there may be no pump seal leak, you don t know what the inside of the pump looks like. It could be eroded and corroded into feebleness. Also, the thermostat might not be opening fully, which may be the easiest to check physically. Thermostats usually stick closed or open, rarely in some midposition that restricts coolant flow. However, we ve seen stats with broken bridges, where the valve opens, cocks and sticks partly open. And if the thermostat has ever been replaced, all bets This new laser gauge is an accurate way to check pulleys for fore & aft and angular alignment. You can pretest the gauge on this cardboard aimer to see if the light hits the correct line. In use, the gauge fits on a ribbed pulley. Using straight-line sighting, the laser beam should hit the approximate midpoint of all other pulleys if they re aligned. are off. The wrong stat definitely can be responsible for partial opening. For example, using a bypass stat in a nonbypass housing can cause the thermostat to hit some part of the housing before it fully opens. A sure cause of poor coolant flow is low coolant level, and today s convoluted cooling system flow patterns can make air pockets a lot more common a cause of this Izzit than a plugged radiator or weak pump. If the cooling system was serviced within some reasonable period (and the radiator was drained), follow the factory procedure for topping it up and air purging it. You can also use an inexpensive aftermarket vacuum assist device that s made for this problem or (even better) a coolant exchanger or coolant flush & fill machine. If the overheating problem seems to come down to the pump or the radiator, your best approach is to gather as much evidence as possible. Warm up the engine and turn on the heater yes, even if it s 90 F in the shade. At fast idle it takes a healthy water pump to deliver a solid flow of hot coolant to the heater. So if you get a quick burst of very hot air from the heater vents, that s a good sign, and you can disconnect a heater hose to double-check the flow. Be aware, though, that many systems have just so-so flow through the heater circuit even with a good water pump, as we find out when the manufacturers install auxiliary electric pumps or rework the hose circuit to improve flow to the heater. You also can visually check for a healthy coolant flow to the radiator if you have a sight glass that can be installed in the upper radiator hose. Although we don t see it sold widely anymore, a sight glass is included with some coolant exchangers, and even may be available separately from some manufacturers. If there s an immovable column of coolant in the sight glass, that casts suspicion on the radiator, of course. Feeling an upper hose for flow may help, but it s not always indicative of much. Using an infrared thermometer to check temperature on both sides of the thermostat, and the radiator core for restrictions, may be more informative. If the coolant sending unit is mounted nearby, its reading may confirm (approximately) the infrared readings. 26 August 2003

4 BATTLING THE COOLING SYSTEM IZZITS If the spring inside the automatic sepentine belt is weak or broken, you should be able to feel weakness or even free play as you move the tensioner back and forth with a wrench, as this photo shows. If the cooling system has been neglected particularly if the reservoir shows signs of a sealer overdose there could be restrictions that slow coolant flow. The system needs a thorough flush & fill. We re amazed at how many times that s enough to cure this Izzit problem, if done properly. Some radiator shops use radiator flow testers, so unless there s still a weak squirt from a heater hose at part-throttle to point to the engine (or water pump), the radiator is where you re more likely to find restrictions. Are you getting the impression that we ve left the water pump as a last resort? If it isn t leaking, then it s the least likely cause of restricted coolant flow. Is It Bad Coolant Or Electrolysis? Electrolysis in a cooling system is a fact of life. If current flowing in the vehicle s electrical system can t all get through the easiest path to a ground, some of it may flow through the coolant, which is somewhat conductive. No electrical grounding is perfect, so some current flow (reflected in a voltage reading in the coolant) can be expected. But is it really a common cause of the premature failure of heat exchangers? This was the hot button question This water pump suffered severe cavitation erosion/corrosion. But because water pumps are not easy to remove in fact, very difficult because they re recessed on many newer engines you don t want to make a mistake by taking one out without a reliable diagnosis. during a cooling system panel discussion at the 2003 convention of the Mobile Air Conditioning Society (MACS) Worldwide. The panelists included representatives of coolant producers, radiator and heater makers and a coolant test lab. Most said they d never seen electrolysis-caused failures, and only a couple said they d ever encountered it. Does that mean electrolysis-caused failure of heat exchangers is so rare you can cross it off the list? Absolutely not. There isn t a lot of research on the subject. However, it s pretty certain that if a radiator or heater core fails within a few thousand miles, or a month or so after replacement, it s probably not bad coolant. Check for electrolysis. Is this also saying the converse is true, that electrolysis is never a cause of, or contributor to, a long-term problem? We haven t seen any definitive research that answers that question, either. There are standard ASTM (American Society for Testing and Materials) tests of cooling system materials for electrolysis. These tests have guided vehicle makers to set.3 to.5 volt as the maximum amount of voltage you should measure in the coolant (voltmeter positive in coolant, negative at battery ground terminal, engine at fast idle for adequate coolant circulation). If the numbers are higher than that, there are one or more weak grounds you should correct. Does that mean that close to.3 volt or so is harmless, even over the long term? Not exactly. What it means is that if the voltage is within a certain range, a cooling system metal should pass a certain ASTM test. Our own voltmeter tests of new cooling systems show voltage readings of under.1 in the best cases to under.2 volt in the worst. Coolant aging apparently has a minor effect on the voltage readings. But there are Gates studies (as well as technicians reports we ve seen) that show excessive voltage may drop significantly after a coolant exchange. That doesn t mean the electrolysis problem is cured because, as noted, poor grounding somewhere caused it, and you have to find and fix that. What if the voltage is above normal and the coolant is bad? Do we get a combination attack on the heat exchangers? Sorry, there are no positive answers to that question. Remember, if a heat exchanger that can be expected to last eight to ten years actually leaks 28 August 2003

5 BATTLING THE COOLING SYSTEM IZZITS Sight glass kits installed in cooling system hoses (arrow) for checking system flow are a useful tool for isolating the Is it the pump or... problems. or fails from corrosion in five or six, it s just going to the metal recycler. It s not going back for analysis to a coolant chemist or vehicle manufacturer s warranty analysis department, or the service department of a manufacturer of heat exchangers. That s real world. This may explain why some of the MACS panelists have not seen electrolysis failures. Before you complain, you should realize that if there were evidence that many heat exchangers were failing in service well before the durability tests say they should, the aftermarket would see that. The carmakers networks of independent shops would be telling them, and the manufacturers then would revisit the test procedures. So even if the tests aren t as definitive as we d like, they seem to have some basis in reality. When a failed exchanger does go back for analysis and the problem is Orange coolant means DexCool to most technicians, but DexCool isn t the only orange coolant. In this case, it s the dye color of the low-silicate coolant used by the Chrysler Group. This photo is from a Dodge truck. recognized as electrolysis that is, high voltage in the coolant causing removal of metal from cooling system surfaces the failure may be plugging or leakage. Heat exchanger manufacturers have reported examples of both, which makes sense. If enough material gets into the coolant and accumulates at small passages, plugging will result. But if the plugging progresses more slowly, the exchanger may fail because of a leak instead. Say you encounter a failure in a heat exchanger and the voltage reading in the coolant is excessive. What do you do? Check all the grounds, of course, and disconnect circuits from the battery, if necessary, to see if the voltage reading drops. One case cited at the MACS convention was a radiator that became corrosion-plugged from electrolysis caused by grounding an aftermarket alarm system at the radiator brace. This provided a ground path through the coolant. The radiator became plugged inside of 3000 miles. In another case, the starter solenoid contacts on a pickup truck weren t separating completely and there was current flow across them. It wasn t enough to complete the starter circuit, but it did result in repeated radiator failure (pinhole leaks and tubes restricted by metallic deposits). Two examples aren t a trend, but they do tell you to be alert. Keep the reverse situation in mind, as well. Say a quick repeat failure indicates a possibility of electrolysis, but the voltage reading in the coolant is normal. Check for a voltage increase as you crank the engine. With the engine running, energize the accessories (including actuating the radiator electric fans) and see if the voltage reading goes above the specs. Poorly grounded fans are a common source of current flow through the coolant. Incidentally, the Gates studies were undertaken because the company had identified electrolysis as the primary cause of internal failure of the upper radiator hose (cracks along the interior length of the hose) in some of their products. Gates reformulated its hoses to reduce their susceptibility to the problem, but noted the condition wasn t 30 August 2003

6 COOLING SYSTEM IZZITS always accompanied by a high current flow through the coolant. The root causes enough to affect hose life include a galvanic loop (current generating battery effect between metal surfaces, past which coolant flows) and the high conductivity of carbon black. However, because voltage readings above the.3 to.5 range seem to be a part of the picture, don t ignore them as a factor affecting hose life. Is It Orange or Yellow? We re talking about the coolant here, and everyone knows that orange is the color of the nonsilicated antifreeze used by General Motors (as DexCool) and at VW/Audi. Yellow is the color of Prestone, and now it s also the color of Zerex G-05, which is used on newer Ford products. We ve said it over and over, but it doesn t seem to get through: You can t tell an antifreeze by color (the photo on page 30 is an example). The reservoir obviously contains orange coolant, but it isn t DexCool or anything like it. It s Zerex G-05, the same stuff used by Ford (and sold in the aftermarket by Valvoline/Zerex) with a yellow dye. And Prestone s yellow (actually more of a gold with a greenish cast) is a conventional American antifreeze, in the same family as the stuff we call American green. But American green and Japanese green are nothing like each other. The bottom line is that all major antifreezes are primarily (about 93% to 95%) ethylene glycol, aside from a few propylene glycol brands sold as safer (less poisonous). So it would seem that if you really flushed a system getting out over 90% of the old coolant could you refill it with any reputable antifreeze? The answer is Yes, with some caveats: Don t just drain the system (even with the lower radiator hose unplugged), because unless you pull the engine and actually turn it upside down, you re not going to remove most of the old coolant. You can drain, top up with pure water, thermocycle (warm up and cool down) the engine, drain again and repeat this procedure a couple more times. That should leave less than 10% of the old coolant in the system. Draining enough water from the system to make room for 50% new antifreeze should be okay. Of course, you ll also have to flush the heater and clean out the reservoir. Sound like too much work? Then get a coolant exchange machine and you ll save a lot of time and effort. Yellow coolant could be anything, possibly even Prestone (which is more of a gold color). But in this case, it s the same stuff that s in the Dodge truck (lowsilicate), except that the photo was taken on a Ford truck. August

7 COOLING SYSTEM If you re installing replacement parts with copper (such as some aftermarket radiators and heater cores), you should know they need protection that may not be provided by some of the new antifreezes that were formulated for the original equipment aluminum era. It s not just copper protection; these antifreezes do contain an inhibitor that addresses that. But if the radiator is assembled with high-lead solder, it needs more specific protection to avoid solder corrosion. Maybe the best idea is to install a premium aluminum radiator or possibly one of the new furnace-brazed copper designs, which are free of highlead solder. If a vehicle is run heavily loaded, particularly in mountainous areas, the engine may be prone to water pump cavitation (where the coolant boils and the bubbles collapse in the pump, eroding the chamber and exposing the surface to corrosion). In that case, an antifreeze with silicates may provide better protection because silicates work faster than other inhibitors. Conventional American antifreezes and G- 05 contain silicates, an inhibitor that provides fast, effective protection of aluminum surfaces. The question of antifreeze life has no simple answer. Sure, the recommendations for conventional American antifreeze range from 2 years/30,000 miles to 4 years/50,000 miles. G-05 is an old heavy-duty formula with a good track record, and has been validated by Ford and Chrysler for 5 years/100,000 miles. GM has approved DexCool for 5 years/150,000 miles. But these recommendations presume a system that 1. hasn t run with a low coolant level, 2. has been topped up with a antifreeze/water mixture and 3. has not used very hard water (if water in your area is very hard, use softened or distilled water). And they do have that five-year limit, which may occur at far less than the peak mileage. Low coolant levels and low antifreeze concentrations (under 40%) are a prescription for corrosion failure. Visit to download a free copy of this article. 36 August 2003