Rattling then smoke

2005 Dodge Caravan - Driving to work this morning. Last mile of a ten-mile trip, there is a steep hill. As I accelerated the car began a rattling sound in the engine compartment. Never heard that before. Also, smelled like something burning. Rattling stopped after topping the hill, but would begin occasionally whenever moderately accelerating, then go away. Oil and temperature gauges read normal.

Got to work, and checked the fluids. At acceptable levels. At lunchtime, decided to try to drive it home. All quiet for the first mile and a half, then rattling started up again. This time, instead of just at moderate acceleration, it also sometimes rattled when coming to a stop or turning. At five miles I had to stop at a store to pick up something. Wasn’t gone long, and when I started it back up and drove off, no rattling until about another mile down the road.

Four miles later, the rattling got worse and did not stop. I pulled into a parking lot just as the engine died and smoke poured out from under the hood. The hood release is broken, so it takes some time to pop it. By the time I got it open, the smoke was gone so I did not see where it was coming from. Could not see any obvious problems. Let it sit for five minutes, then it started right up. No noises, and couldn’t see any issues while running.

Less than a mile from home, so I decided to go ahead and drive it. The van behaved until about two blocks from the house, then the occasional rattling began again. Made it into the driveway.

So, where do I begin troubleshooting? Thanks!

Does the rattling sound like a loose rod or more like heavy detonation i.e. pinging?

“Got to work, and checked the fluids. At acceptable levels”

If I am to interpret that to mean that you checked the fluids right after arriving at work, then I have to assume that you didn’t open the radiator cap, as you would have likely been bathed in scalding liquid. So…I am going to suggest that you open the radiator cap when the engine is stone-cold, and report back to us on what you found.

My theory–which, of course, could be totally wrong–is that you may have seen steam, rather than smoke, and that your engine may have been overheating. When an engine is overheating, there can be some…unusual…noises heard as the resultant damage takes place. If the engine was overheating, at the very least, you may have been hearing the engine “knock” as a result of high temperatures.

Please report back to us on what you find when you open the radiator cap.
Also…are you using the correct 50/50 coolant mix in the cooling system, instead of just water?

Hearing unusual noises from the engine compartment when going up steep hills, the first thing I think of is engine pinging. That’s the condition when pinging will be noticed, if the engine is on the verge of pinging. Of course before doing anything else, once everything cools down, check the oil and transmission fluid and the coolant levels to make sure they are good. For the coolant it isn’t sufficient to just look at the plastic bottle, you have to remove the cap and peer into the radiator too.

Pinging can be caused when the spark plugs are left in the engine too many miles and the gap opens up. What’s the routine maintenance situation? Any engine stuff lagging behind the schedule the manufacturer recommends?

I tend to second VDCdriver and I’m also curious about the engine coolant level. A full coolant recovery tank doesn’t necessarily mean the radiator itself is full.

Maybe a thermostat intermittently sticking and causing overheating?

Maybe a water pump getting a little noisy?

Offhand, I’m leaning towards an engine coolant issue of some sort although an engine suffering severe pre-ignition rattles can overheat. Overheating can also cause that rattle as VDCdriver mentioned.

It’s also possible for a temp gauge to show normal even while low on coolant if the temp sender probe is not immersed in hot coolant; all depending upon engine type.

“It’s also possible for a temp gauge to show normal even while low on coolant if the temp sender probe is not immersed in hot coolant”

I can recall a number of engines that were overheating, even though the temp gauge didn’t indicate a high temperature. In every case, they were very low on coolant.

If a vehicle has an actual gauge & not an idiot light wouldn’t the gauge read a lower than normal temperature reading if the sensor wasn’t immersed in hot coolant ? Providing the gauge is working properly ?

...wouldn't the gauge read a lower than normal temperature reading if the sensor wasn't immersed in hot coolant ?

Not necessarily. It depends on the way the sensor is mounted; how fast heat is transferred out of the sensor.

Sometimes it may be related to the vacuum system, the vacuum canister (or parts/valve) and filter. This area is often overlooked and many times it’s a minor problem (especially if there are no check engine lights illuminated). Good luck!

Darren, I suspect that after two years the problem has been solved. :grin: