# Why water cracks hot exhaust manifold

1. May 27, 2007

### honestrosewater

This cute mechanical engineer mentioned in a message that he had designed some manifolds, so that's why I'm asking this (in addition to my general interest in physics, of course).

I'm not sure if they were intake or exhaust manifolds, but that's not the problem now. I figured that http://en.wikipedia.org/wiki/Manifold_(automotive_engineering)" [Broken] was a good place to start, and it says
I want to figure out why that would happen. I'd really like to see if I can figure it out on my own with a little bit of guidance.

It says that exhaust manifolds are usually made of cast iron. Do I need to look up what types of forces and bonds are at work in cast iron's solid structure?

I'm guessing that the liquid water is cooler than the iron, so it will cool the iron where it drips. I can't think of a reason why this alone would cause the iron to crack. Is this enough?

Does some chemical reaction take place first, and then it cracks due to some mechanical forces, perhaps because the new chemical structure is weaker?

I'm basically just looking for a little direction since I'm curious but don't have tons of time to try every path.

Last edited by a moderator: May 2, 2017
2. May 28, 2007

### siddharth

Exactly. So, do you know how the material properties of objects, such as the length or shape depends on temperature?

3. May 28, 2007

### honestrosewater

Maybe. At a higher (macroscopic?) scale, heat would make a material expand, I think. The temperature of a system is the average kinetic energy of the particles in the system, is that right? So increasing or decreasing the temperature is accelerating or decelerating the particles, respectively? The faster that the particles are moving, the more often they collide and push each other apart? I should stop there before I wander too far in the wrong direction. I think that I am thinking more of gases, and I'm not sure if it would work the same for solids or if there are additional things that I would need to consider.

Also, I'm not sure what's going on at the intermolecular level, um, in terms of the electromagnetic forces that are holding the molecules together. (I assume they are electromagnetic forces. Yes?)

Oh, and cooling it would cause it to contract. So if that's right, I need to figure out why that would cause it to crack. I'm not sure what type of failure cracking is. Haha, do you know what I mean? I will think on it.

Oh, I have been thinking that the result would be a single crack. But would it perhaps instead cause two or more cracks, along the edges of where the water was dripping, i.e., where the cooler material pulled away as it shrank from the hotter surrounding material?

Last edited: May 28, 2007
4. May 28, 2007

### siddharth

Yeah. This tendency is called as http://en.wikipedia.org/wiki/Thermal_expansion" [Broken].

Yeah. It's the measure of the average kinetic energy of the particles in a system.

That's right. For a gas, the molecules are far apart and don't interact except during collisions. So, as you increase the temperature, the molecules move about faster on average, and collide with increased frequency.

Right again. In solids, the molecules are closely held (ie, bonded) by electromagnetic forces and can't move around as much. These molecule, which are held together by the electromagnetic forces, are similar to springs. As you increase the temperature, kinetic energy increases through vibration of the molecules, and the molecules move further apart as the bonds weaken.

Are you familiar with the concept of stress and strain in materials? As far as I know, while the general principle of failure due to thermal stress can be understood, I think that the exact mechanism by which it fails is a complex phenomenon.

So, I don't know what type of failure cracking is. Maybe, it's the result of the cyclic expansion and contraction as the water drips.

Last edited by a moderator: May 2, 2017
5. May 28, 2007

### honestrosewater

Alas, no. But I can fix that.

Cool, those are good leads. Thanks bunches!

6. May 28, 2007

### Danger

Rosie, you have one great mind. You pretty much answered all of your own questions.
Generally, it's simply a matter of the quenched area contracting while the hot part doesn't, which essentially pulls them apart. There is another effect which I've witnessed, but I don't know if it's applicable in this case. It might in fact have been a single fluke incident that will never happen again. Cast iron is somewhat porous. I saw an instance where water infiltrated a minor crack, and then explosively widened it when it turned into steam. Mind you, that was a cheap Hibachi, not a manifold.

7. May 28, 2007

### honestrosewater

It struck me as something that I might know enough to answer, so maybe that is to be expected.

Is that what happens then, you end up with cracks along the edges of the water-contact area rather than cracking along the center of this area? I can understand the edges case, methinks. It is a tensile force, no, where the colder material is pulling on the hotter material, or along the temperature gradient? Or, hm, I'm not sure how much I know about stress and strain. I should probably read more first.

8. May 28, 2007

### Astronuc

Staff Emeritus
Cracking is basically the separation of planes of atoms in a solid. It does occur under tensile (force normal to plane of atoms) or shear (transverse displacement) loads/stress.

As one has correctly surmised, a cool region next to much hotter region will experience differential thermal expansion/contraction. The cooler region tries to contract but is contrained by the hot region which has expanded. The cooler region then experiences a tensile stress, and a crack will likely form near the location of maximum tensile stress.

Cracks tend to form near discontinuities in the microstructure, e.g. different phases, inclusions or flaws. Flaws are essentially microcracks waiting to grow.

9. May 28, 2007

### honestrosewater

Ohhh... So I'm going to have to learn a bit about tensors in order to get a clear picture of the forces here. Check.
In this case of solid cast iron, the thermal energy is being transfered by conduction, yes? Does thermal radiation occur only at surfaces? To me, this means that the surface atoms were excited and emitted photons. It seems odd then to think of conduction as not being also the emission and absorption of photons in some sense, but no one seems to explain it this way. But then I read about phonons. Are phonons like photons in some way (except not being gauge bosons)? Am I wrong in inferring that the surface atoms were excited? How many ways are there to excite an atom? I can let that go if it's too complicated since I don't really know enough to take it much further yet. But I'm quite curious.

I looked at some pictures of cracked manifolds, and it's not what I was thinking. There is usually a single crack. But now I wonder whether such a crack is necessarily the result of tensile stress on the cooler surface, i.e., whether the location of the crack tell you where the drip was, or could the crack have resulted from a process like my next question...

I'm reading more to figure out exactly how this might happen if it can, but could it happen that sending heat "waves" through the solids magnifies, or worsens, the inconsistencies that are already present, perhaps by making the waves interfere with themselves in some way? I'm perhaps on thin ice calling them waves to being with, but I'm wondering if that's what Astronuc was getting at with the last comment.

Is the former, the cooler region cracking, a case of thermal shock?

Hah, it didn't take long for me to get in over my head. :tongue2: Sorry. I am still working on answering them by myself too.

Last edited: May 28, 2007
10. May 28, 2007

### FredGarvin

Although in the general loading case, the stress field is described by a tensor, that is not what Astro meant by tensile. Tensile simply means a normal stress induced by pulling in opposite directions, i.e. a tensile test.

In my experience, there are many cases that induce cracks. Each scenario presented so far are applicable. I won't get into photon emissions or the like, but you have three basic options to consider:

1) The thermal stresses were the culprit and essentially tore the thing apart.
2) An occlusion helped in starting a crack and the thermal stresses just got it started.
3) If there was enough water, the local area could be rapidly quenched causing an area of increased hardness and brittleness.

The best way to determine this is to take samples and have them etched so you can look at them in a lab.

11. May 28, 2007

### Astronuc

Staff Emeritus
:rofl: Curiosity is not fatal, and in fact it is a good attribute to have. Lots of questions - which are not difficult to answer, but altogether it would take a while to answer satisfactorily.

In the case of solids (and liquids), thermal energy is transferred via conduction. Liquids can experience convection under the right conditions, and liquids and solids can radiate energy, although liquids tend to evaporate (change of phase from liquid to vapor), although some solids can sublime (solid to liquid phase change.) In the case of a case iron manifold, the heat is radiated from the surface, but that is a small fraction of the heat transfer since the manifold is just not very hot for radiating heat efficiently. Most of the heat will be conducted to the air outside the manifold, and some convection will take place - cooler air displacing the warmer/hotter air.

In the case of a water dripping on the manifold, the water flashes to steam (liquid to vapor phase change) and the rapid cooling sets up small area of 'thermal shock'. A rapid quenck produces a localized cooling, and that area tries to shrink, but can't because the hotter sections constrain it. Internally, somewhere in the cold region are high local tensile stresses (tensor mechanics can be used to describe the stress field).

Phonons and photons are not the same, nor really similar. Phonons are related to atomic/molecular vibration, and are related to heat conduction in solids and liquids (as well as sound conduction). Photons are electromagetic phenomena. Energy can be transferred by phonons (conduction) and EM radiation (as in blackbody radiation).

HRW, you're not wrong in assuming surface atoms are excited - they are, but so are all the atoms in something that is heated. Surface atoms can radiate heat via EM radiation (photons), as well as conduct heat to cooler atoms just below the surface, or to gas or liquid atoms or molecules in contact with the surface atoms.

12. May 28, 2007

### turbo

One key concept HRW - when water impinges on a hot exhaust manifold, it flashes to steam, and the amount of heat energy carried away by the latent heat of vaporization is enormous. The water is not just cooling the manifold as the water warms, it is providing a heat-sink that causes rapid localized cooling as the iron gives up heat to flash the steam. The latent heat of vaporization is a tremendous "leverage" in thermodynamics and capturing and releasing this heat is the key in steam-turbine power plants.

13. May 29, 2007

### honestrosewater

Haha, physics is so cool. Thanks, you guys are super.

I don't really get it all yet, but I've made some nice progress. I am relatively comfortable already with electromagnetic radiation and atomic structure. The things that I need to work on are heat and the intermolecular forces in different phases. I might end up with some coherent questions down the road, but there is one question that I think might really help:

Would it be consistent to define thermal conduction and convection as the transfer of thermal energy via virtual photons and define thermal radiation as the transfer of thermal energy via real photons?

Even as I say it, something doesn't make sense. Isn't the whole idea of thermal radiation assuming that these photons are in some sense making "thermal contact" (or they are the thermal contact) with some other (cooler) receiving system whose temperature differs from the source system? What is the other system? Its environment? Oh, haha, I guess that would make sense.

14. Sep 14, 2007

### kuba

I guess all the pieces are there, but you're overcomplicating things a bit. Water drops are falling on a cast-iron exhaust manifold. The manifold will be hot - several hundred C (lets say). So, as has been said, the water will flash into steam, so it will take quite a bit of heat with it. Now, as this is a transient process, likely much faster than heat conduction through the manifold, the metal will locally cool down.

Then you have to understand why the manifold will crack: due to stresses (i.e. internal forces), and some material discontinuity which would be a starting point. Cast irons have inherent discontinuities built into them, by design, as far as I recall.

You can induce stresses simply by mechanically loading the material -- bang the manifold with a sharply pointed, surface-hardened hammer and it will likely crack. Another way of inducing stresses is to locally change the temperature of the material. As it's cooled down, it will locally contract, but it's still bound to the rest of the hot manifold. What happens is that the hot manifold (surrounding material) will stretch (strain) the now-contracted piece of the cooled-down manifold, such that it remains a contiguous piece. If the stress required to keep the cooled-down spot stretched to the heated-up size is big enough, it may initiate cracking. Cracking is investigated by fracture mechanics and is pretty interesting. Suffice it to say that if the stresses at the tip of the crack are big enough, the crack will keep on growing.

This is it, in a nutshell.

15. Aug 16, 2009

### Tony Rest

Continuing honestrosewater's line of thought, I am reminded of what happens when a hot ceramic or glass container is placed in a refrigerator: a cracking occurs. In this instance the contracting exterior (cold side) pulls away from the still hot expanding interior of the vessel. The macro effect of the of the pulling away at the hot-cold juncture eventuates in a series progressive separation cracks working their way towards the interior of the vessel as deeper strata are exposed to the cold exterior environment. In the manifold instance the contact point of the water induces a pulling away from the hot expanded molecule perimeter towards the cooler center of the the water's contact point. Both are radial phenomenon, but in the refrigerator instance the heated molecules are in the center and the cold ones at the perimeter.

16. Aug 17, 2009

### doBBin

Hey all, first post!! I'm not quite sure whether this is a valid statement (perhaps because I am a 14 year old) but wouldn't it be a simple matter of thermodynamics; due to the fact that the water and the super heated iron exchange heat rapidly with the water, causing the iron to contract. For example if you take a strong cast steel saucepan and heat it to a very hot temperature, then pour room temperature water on it, the pot will crack, as the metal contracts so quickly in the area it has been touched. It cracks in the contact area as the surrounding metal is still hot. This is similar to a hot exhaust manifold made of even weaker cast iron which has several imperfections within its crystalline structure. If heated to a temperature of say 800 degrees Celsius it is even more likely to crack if water was to touch the heated iron the metal would contract causing a small fracture.

If the iron was heated over and over and over the metal would become tempered making it stiffer and more brittle making it even more vulnerable to cracking.

Last edited: Aug 17, 2009