Wasted heat from combustion engines...

[TurtleCrusher]

Why isn't the wasted heat given off used for anything besides interior heating inside of a vehicle? Is there a way to harness that heat and turn it into electricity?

Probably a dumb question...

 

[wuliheron]

Today's cars are so efficient at burning gas that for your average 20 mile commute roughly half the gas and most of the pollution are created in the first few minutes due to the engine not being hot enough yet. However, some of the newer cars use the exhaust to preheat the fuel for better combustion. The issue then is not just how to use wast heat, but how to do so selectively.

The first principle of recycling is its always best to reuse whatever it is for the original purpose for which it was made. In this case, heat is created to provide torque to turn the engine. If engines could be made of plastic or something else that didn't conduct heat quite so well then it would be easier to recycle the heat right where it would do the most good with the minimum amount of loss. It may sound crazy, but there is at least one such attempt that I know of and the weight savings is substantial as well.

Capturing just some of the heat as it leaves the engine and converting it to electricity then is not the best choice. It adds extra parts and weight meaning additional costs that must justify themselves. The cost and weight of parts to convert heat into electricity is low, but so is their efficiency making them not yet attractive to auto makers.

 

[Evadman]

Sterling engines could probably be used to harness additional power from the wasted heat if they could be made light enough.

 

[Mark R]

Modern cars are very efficient at burning gas. The problem is that small reciprocating engines aren't very efficient. Gas is worse than diesel, because the throttle losses are substantial. There's nothing useful that can be done about adiabtic heating losses - although, perhaps, exotic thermally insulating ceramics could be used for key engine parts. However, they can be improved on in other ways e.g. by redesigning key engine concepts. E.g. an Atkinson cycle engine is more efficient than an Otto cycle (conventional gas) because it has asymmetric heating/cooling phases. The prius uses an atkinson engine, which is contributes a large part of its improved fuel efficiency.

The other issue is frictional losses - there are a large number of bearing or friction surfaces in a modern engine, and while there have been advances in materials, specialist honing techniques and lubricants, there isn't much scope for improvement. Thanks to modern multi-grade lubricants there is little need for a 'warm-up' period to get optimal lubrication.

Of course, some technology is retrograde. Catalytic converters are now mandated to prevent carbon monoxide and nitrogen oxides emissions. The problem is that modern engines combust so efficiently, that the catalytic converter may not be able to reach operating temperature without assistance. Because of that, ECUs must be programmed to degrade engine efficiency while the catalyst is cold, in order that the catalyst can warm up fully.

To harness heat, you need a heat engine. The 2nd law of thermodynamics limits the maximum amount of energy a heat engine can extract - this is called the Carnot efficiency, and depends on the temperatures of the 'hot' and 'cold' sides of the engine. The larger the temperature difference, the more useful the heat, as the more efficiently it can be harnessed. A car engine works well, because it has uses 'cold' air at room temp, and 'hot' air at combustion temp (1000 C +). However, a heat recovery system connected to the engine coolant would be worthless, because coolant at 90 C and air at 20 C doesn't provide enough difference to be captured with a worthwhile efficiency.

This means the only realistic source of energy capture is the exhaust - this is already done in some vehicles by the use of turbochargers. However, there are considerable engineering issues with the use of such high-speed turbomachinery in such a high temperature, corrosive environment. However, a number of groups have developed turbo-alternators which replace the conventional pulley alternator. Unfortunately, the brushless DC generators required for such high-speed operation are complex, microprocessor controlled devices, quite unlike a conventional alternator.

 

[Aluvus]

Is there a way to harness that heat and turn it into electricity?

In principle, you can place a Peltier device on whatever (non-moving) hot surface is convenient. Peltier devices generate electrical current when there is a temperature difference between their two surfaces. However, the amount of power you can generate at reasonable temperatures is just not very large.

 

[pcgeek11]

Today's cars are so efficient at burning gas that for your average 20 mile commute roughly half the gas and most of the pollution are created in the first few minutes due to the engine not being hot enough yet. However, some of the newer cars use the exhaust to preheat the fuel for better combustion.

They did the same thing in the 50'sand 60's by using the exhaust manifold heat to warm the intake air.

 

[wirednuts]

Of course, some technology is retrograde. Catalytic converters are now mandated to prevent carbon monoxide and nitrogen oxides emissions. The problem is that modern engines combust so efficiently, that the catalytic converter may not be able to reach operating temperature without assistance. Because of that, ECUs must be programmed to degrade engine efficiency while the catalyst is cold, in order that the catalyst can warm up fully.
.

basically, what you are referring to is "open loop" mode. aka, dummy mode. when the engine is cold, the o2 sensors cannot get accurate readings so the ecu just feeds a rich mixture to the engine to keep it running. degrading engine efficiency wont help it heat up faster though... the only reason they run a rich mixture in open loop mode is because its more reliable to have a mixture too rich over one too lean.

i want to add, the flipside to this topic in the OP, is about electric cars. they dont have extra heat, so if they NEED heat in the winter, you have to suck down battery juice. HUGE limitation. it will be a long time before electric cars make mainstream in the north states.

 

[Matt1970]

If anything combustion effiency will warm an engine up quicker.

 

[JimW1949]

I want to add, the flipside to this topic in the OP, is about electric cars. they dont have extra heat, so if they NEED heat in the winter, you have to suck down battery juice. HUGE limitation. it will be a long time before electric cars make mainstream in the north states.

I agree with you, heating an electric car by using the batteries would not be a good idea due to the high drain on the batteries. While it is certainly true you would need at least SOME heat to melt ice from the windshield, and to make the car somewhat comfortable to ride in, why does the heat for the car have to come from the batteries? Why can't you have a propane tank and a small heating unit to supply heat to the vehicle? You could even use the propane to heat the battery compartment. If the battery compartment were well insulated I doubt it would take much propane to heat it and keep it warm.

 

[Lemon law]

First of all, when the engine is cold, often below 0F, heating the fuel does little, because when the vaporized fuel hits cold engine parts, it condenses back into a liquid. Of course when cold engine parts warm up to 150+F, that effect dramatically reduces. But meanwhile, the gas engine needs a far richer gas mixture to compensate.

The second thing to note is that gasoline engines are at best 20% efficient. Some is lost to inefficiency in gearing and the like, but the bulk is lost in waste heat. Any way to recover that lost waste heat is not yet really practical, but still 80% inefficiency is really bad. Diesel engines are better at 30%, but still are very inefficient also.

The other things to note is that gas electric hybrids are great for mixed city traffic jam
situations, but would suck for my mainly only highways speed driving situation.

 

[pandemonium]

I think our implementation of energy conversion is very poor in general.

Take nuclear power plants for example. All we're doing is generating heat off of nuclear fission to boil water and make steam to rotate turbines? That concept is extremely lackluster for today's age.

I think the fact that we're focusing on alternative energy means instead of making use of the energy at our disposal is some real backwards logic. In some ways "alternative" does mean a better conversion method, but most of the time it's just another poor performer in the world of energy utilization.

 

[Farmer]

Take nuclear power plants for example. All we're doing is generating heat off of nuclear fission to boil water and make steam to rotate turbines? That concept is extremely lackluster for today's age.

That would require some kind of scientific breakthrough.

It might sound quite antiquated, since the Rankine cycle has been around since the industrial revolution, but that is the just the concept, not the implementation. Improvement comes from removing losses (pump losses, turbine losses, losses from viscous forces on the working fluid) and increasing the temperature difference between the hot (reactor) and cold (condenser) reservoirs as to increase the maximum possible efficiency of the cycle. This becomes a problem since increasing the reactor temperature means increasing the turbine inlet temperature. Both aims are at the cutting edge of materials science, as you want materials that steel-like in material properties, but with much greater temperature tolerance.

A similar concern is found in modern jet engine design: you want a hot combustion chamber, but you don't want to melt or warp your turbine rotors. Does that sound more high tech to you?

In another way, classical electrodynamics was solved in the mid 1800s. It doesn't mean that current technology based on this theory, like hybrid or electric vehicles, is "lackluster."

 

[pandemonium]

That would require some kind of scientific breakthrough.

It might sound quite antiquated, since the Rankine cycle has been around since the industrial revolution, but that is the just the concept, not the implementation. Improvement comes from removing losses (pump losses, turbine losses, losses from viscous forces on the working fluid) and increasing the temperature difference between the hot (reactor) and cold (condenser) reservoirs as to increase the maximum possible efficiency of the cycle. This becomes a problem since increasing the reactor temperature means increasing the turbine inlet temperature. Both aims are at the cutting edge of materials science, as you want materials that steel-like in material properties, but with much greater temperature tolerance.

My problem is with our focus on how we're converting the energy in such an energetic phenomenom as nuclear fission. Instead of utilizing the kinetic energy directly, such as that of a combustion engine or jet engine, we're simply harvesting the by-product that's produced: heat. Am I wrong? Is heat not the by-product? I have a difficult time understanding how such is the case if in chemical reactions the by-product is more-less the bane of the process rather than the product that we're aiming for.

I would think we can utilize other methods of energy conversion as well as the Rankine method at nuclear sites. I also realize that creating a "nuclear engine" much like that of the typical combustion engine, is quite a bit more involved. This is where I'm stating where our focus should be - utilizing more than just the thermal by-product of the reaction.

I may be crazy to picture giant pistons laid out on their side or at a 45 degree angle on slides driven by a nuclear reaction chamber that pumps dynamos at the maximum rate of 3 RPM (as far as I know our conversion rates from kinetic to electrical is fairly adequate) and have the coolant system pumping the heat exchange from the process to power turbines the same as it does today.

I'm sure once the math is in place and tested properly if you get the right size, weight and design of 4 pistons (maybe in a semi criss-cross pattern) we could find that method of kinetic energy conversion for nuclear fission.

*shrugs* I'm just a guy that likes to ponder. I have no formal knowledge of such things.

 

[imagoon]

My problem is with our focus on how we're converting the energy in such an energetic phenomenom as nuclear fission. Instead of utilizing the kinetic energy directly, such as that of a combustion engine or jet engine, we're simply harvesting the by-product that's produced: heat. Am I wrong? Is heat not the by-product? I have a difficult time understanding how such is the case if in chemical reactions the by-product is more-less the bane of the process rather than the product that we're aiming for.

I would think we can utilize other methods of energy conversion as well as the Rankine method at nuclear sites. I also realize that creating a "nuclear engine" much like that of the typical combustion engine, is quite a bit more involved. This is where I'm stating where our focus should be - utilizing more than just the thermal by-product of the reaction.

I may be crazy to picture giant pistons laid out on their side or at a 45 degree angle on slides driven by a nuclear reaction chamber that pumps dynamos at the maximum rate of 3 RPM (as far as I know our conversion rates from kinetic to electrical is fairly adequate) and have the coolant system pumping the heat exchange from the process to power turbines the same as it does today.

I'm sure once the math is in place and tested properly if you get the right size, weight and design of 4 pistons (maybe in a semi criss-cross pattern) we could find that method of kinetic energy conversion for nuclear fission.

*shrugs* I'm just a guy that likes to ponder. I have no formal knowledge of such things.

From my minimal nuclear physics knowledge...

Uranium 238 is basically an exothermic decay process. The Alpha (hydrogen) and beta (electrons) expelling from the nucleus produce minimal "thrust." It is the heating and conversion of liquid water to steam that actually produces the movement that drives the turbines.

--edit--

Actually that is not all the different than a combustion engine. Fuel ignites and a) heats the air causing expansion b) the resulting products (CO, CO2 etc) are lower density and propel a piston to turn the engine. A jet turbine is even closer to a reactor in design.

Decay chain:

U-238  	Uranium-238  	alpha  	4,460,000,000 years  	Th-234
Th-234 	Thorium-234 	beta 	24.1 days 	Pa-234
Pa-234 	Protactinium-234 	beta 	1.17 minutes 	U-234
U-234 	Uranium-234 	alpha 	247,000 years 	Th-230
Th-230 	Thorium-230 	alpha 	80,000 years 	Ra-226
Ra-226 	Radium-226 	alpha 	1,602 years 	Rn-222
Rn-222 	Radon-222 	alpha 	3.82 days 	Po-218
Po-218 	Polonium-218 	alpha 	3.05 minutes 	Pb-214
Pb-214 	Lead-214 	beta 	27 minutes 	Bi-214
Bi-214 	Bismuth-214 	beta 	19.7 minutes 	Po-214
Po-214 	Polonium-214 	alpha 	1 microsecond 	Pb-210
Pb-210 	Lead-210 	beta 	22.3 years 	Bi-210
Bi-210 	Bismuth-210 	beta 	5.01 days 	Po-210
Po-210 	Polonium-210 	alpha 	138.4 days 	Pb-206
Pb-206 	Lead-206 	none 	stable 	(none)

 

[Mark R]

My problem is with our focus on how we're converting the energy in such an energetic phenomenom as nuclear fission. Instead of utilizing the kinetic energy directly, such as that of a combustion engine or jet engine, we're simply harvesting the by-product that's produced: heat. Am I wrong? Is heat not the by-product? I have a difficult time understanding how such is the case if in chemical reactions the by-product is more-less the bane of the process rather than the product that we're aiming for.

But that's the point. There is no 'direct' kinetic energy - the kinetic energy is from combustion is random, not directional. There must necessarily be losses when making it directional, and the size of those losses is dependent on the temperature gradient from 'hot side' of the engine to the 'cold side'.

An internal combustion engine is a heat engine, just like a steam turbine, and a jet engine. They follow the same 2nd law of thermodynamics that prescribes the maximum efficiency - beause they are, the same thing - just slightly different ways of doing it.

The complex mechanical systems is an internal combustion engine severely hamper the efficiency. Similarly, jet engines due to weight considerations, may not always operate at the very high temperatures and pressures needed to achieve best efficiency. Although, advanced material designs have made it possible to boost jet engine efficiency, to the point that it is almost as good as piston/propeller engines.

If you scale up a diesel engine to small power-station size (100,000 hp), then you get efficiencies better than 50% - due to improved control of friction, and reduced adiabatic heat losses to the engine substrate.

State of the art, super-temperature gas turbines, can achieve about 45% efficiency. Which can be increased to about 55%, if the exhaust is used to heat water for a steam turbine.

The reason current power stations tend to use water/steam is that the cycle is well understood, and materials for operating at those pressures and temperatures are well characterised. Jet engines, and gas turbines, which operate at extremely high temperatures, require exotic single-crystal superalloys, which cost 20x their weight in silver.

The advantage of water/steam is that the conversion from liquid to gas embodies a lot of energy. Using a gas (e.g. air or helium) would require pumping a much larger mass of gas, to transfer the same amount of energy (because you don't get the 'concentrated' energy of the phase change).

The disadvantage of water/steam is that it becomes difficult to design a good boiler for very high temperatures (> 400C) because steam is much less dense than water (and so less effective at transferring heat from the boiler). This can be worked around by increasing the pressure to such a high level that the steam reaches the same density of water (the water and steam under such conditions act indistinguishably from each other and simply become a single fluid - this is called a 'supercritical' fluid). Modern coal plants use this technology, and it is well understood. The problem with going to supercritical water is that it's viciously corrosive (this is well understood in conventional plants, but supercritical water nuclear plants remain a long way off, as the best known supercritical water compatible materials are not nuclear compatible - research is continuing).

Similarly, there's no reason why you can't take a nuclear heat source, and use it to heat air , in a gas turbine, in exactly the same way as a jet engine heats air in it's turbine. The issue is purely one of materials. At the low temperatures safely achieveable by current nuclear fuel, steam is an excellent method of energy conversion, and is more practical than a jet engine type design (a Brayton cycle engine, to use the technical term).

If we had a nuclear fuel technology that could safely operate at 900-1000 C while gas cooled - then a Brayton cycle would be a fabulous idea.

 

[JimW1949]

My problem is with our focus on how we're converting the energy in such an energetic phenomenom as nuclear fission. Instead of utilizing the kinetic energy directly, such as that of a combustion engine or jet engine, we're simply harvesting the by-product that's produced: heat. Am I wrong? Is heat not the by-product? I have a difficult time understanding how such is the case if in chemical reactions the by-product is more-less the bane of the process rather than the product that we're aiming for.

I like your idea of an engine running on nuclear fission, but I have my doubts it could be done. Maybe I am an eternal pessimist, but I just don't see an engine running on nuclear fission any time in the near future. You may be right, it could happen, but I doubt we will see it in our lifetime.

Generally speaking, nuclear fission is used to build devices in order to destroy things, like cities. The four things normally associated with nuclear fission is heat, light, blast and radiation. Although blast is associated with superheated air rushing outward, then back inward, most people consider heat and blast as two different things.

I have no idea how you could build an engine utilizing the incredible energy released by nuclear fission. I am not saying it isn't possible, I am just saying it is way past our current level of technological expertise. About the only way we can utilize nuclear fission for constructive purposes with our present technology is the way we are currently doing it.

I am sure fusion power would be a much better route to take, but fusion power is way beyond our ability to control it. We have fusion power in a sense, we have hydrogen bombs, but hydrogen bombs is not a controlled way to utilize the enormous energy obtained from joining hydrogen atoms. In fact, it is my understanding that a nuclear fission bomb is used to create the necessary heat in order to start the atoms of hydrogen bonding together.

I don't have any idea what the definitive answer is, all I do know is I am not a huge fan of nuclear power plants because of the dangers associated with them. They are wonderful when they are working correctly, they produce an enormous amount to electricity which we desperately need, but if anything unforeseen happens the consequences can be devastating.

 

[pandemonium]

Actually that is not all the different than a combustion engine. Fuel ignites and a) heats the air causing expansion b) the resulting products (CO, CO2 etc) are lower density and propel a piston to turn the engine. A jet turbine is even closer to a reactor in design.

This is more what I was getting at by mentioning kinetic energy since combustion engines detonate and expand the air (on the molecular level, it would be considered "thrust" in general terms).

I guess one of the differences is that nuclear fission is on the atomic level instead of the molecular level with combustion.

Then again, we have the controlled reactor fission, and the energetic uncontrolled holy mother of destruction "bomb". We can't find a happy medium?

But that's the point. There is no 'direct' kinetic energy - the kinetic energy is from combustion is random, not directional. There must necessarily be losses when making it directional, and the size of those losses is dependent on the temperature gradient from 'hot side' of the engine to the 'cold side'.

An internal combustion engine is a heat engine, just like a steam turbine, and a jet engine. They follow the same 2nd law of thermodynamics that prescribes the maximum efficiency - beause they are, the same thing - just slightly different ways of doing it.

Semantics. As I said above the kinetic thrust is ideally represented by the heat expansion of the gases in combustion and kinect energy released by the Coulomb repulsion in nuclear fission.

From wiki:

When a uranium nucleus fissions into two daughter nuclei fragments, an energy of ~200 MeV is released. For uranium-235 (total mean fission energy 202.5 MeV), typically ~169 MeV appears as the kinetic energy of the daughter nuclei, which fly apart at about 3% of the speed of light, due to Coulomb repulsion. Also, an average of 2.5 neutrons are emitted with a kinetic energy of ~2 MeV each (total of 4.8 MeV). The fission reaction also releases ~7 MeV in prompt gamma ray photons. The latter figure means that a nuclear fission explosion or criticality accident emits about 3.5% of its energy as gamma rays, less than 2.5% of its energy as fast neutrons (total ~ 6%), and the rest as kinetic energy of fission fragments ("heat").

From this I'm inclined to say the kinetic forces are there. We just don't know how to properly manipulate them yet.

I need to sit down with a nuclear physicist and pick their brain. Oddly enough I know an engineer that works at the nuclear facility near. I'll have to discuss this with him.

 

[Weenoman]

An atomic powered engine would be awesome, but just how much more efficient than our current technique would it be?

 

[Howard]

An atomic powered engine would be awesome, but just how much more efficient than our current technique would it be?

Efficiency wouldn't matter; you'd get so much power out of it, you wouldn't know what to do with it.

 

[imagoon]

From this I'm inclined to say the kinetic forces are there. We just don't know how to properly manipulate them yet.

Actually I would think that it already does. The mass of an atom is very small so there is minimal inertia to keep it moving. The kinetic component breaks down in to heat increasing the temperature of the "medium" (water / sodium / etc) which then uses that heat to develop it back in to kinetic energy.

Also 200 MeV is not a lot of power.

http://www.tpub.com/content/doe/h1019v1/css/h1019v1_124.htm

You would need 3.12 x 10^10 fissions to give you 1 watt in one second.

You also cannot know which vector the atoms will take so you cannot guarantee if the kinetic portion would ever make it to your turbine or which ever. The atoms may hit the wall of your reactor and bounce releasing heat etc.

 

[Matt1970]

I wouldn't trust a nuclear fission engine. I remember one of the Bikini island test blasts was only supposed to be 4 or 5 megatons and ended up being 13.

Edit: I looked it up. Castle Bravo March 1, 1954 at Bikini Atoll, Marshall Islands, as the first test of Operation Castle. Castle Bravo was the most powerful nuclear device ever detonated by the United States, with a yield of 15 megatons. That yield, far exceeding the expected yield of 4 to 6 megatons, combined with other factors, led to the most significant accidental radiological contamination ever caused by the United States.

 

[jhu]

There was the proposed project Orion. Never got off the ground though, er..., not funded.

 

[pandemonium]

Actually I would think that it already does. The mass of an atom is very small so there is minimal inertia to keep it moving. The kinetic component breaks down in to heat increasing the temperature of the "medium" (water / sodium / etc) which then uses that heat to develop it back in to kinetic energy.

Also 200 MeV is not a lot of power.

I'm not sure that's entirely accurate, seeing how everywhere I see mention of comparisons between nuclear reactions and combustion oxidation the output energy is much, much higher with nuclear fission. The paragraph above the one I quoted earlier mentions this. I was trying to put it into math but work got in the way. >.<

Typical fission events release about two hundred million eV (200 MeV) of energy for each fission event. By contrast, most chemicaloxidation reactions (such as burning coal or TNT) release at most a few eV per event, so nuclear fuel contains at least ten million times more usable energy per unit mass than does chemical fuel. The energy of nuclear fission is released as kinetic energy of the fission products and fragments, and as electromagnetic radiation in the form of gamma rays; in a nuclear reactor, the energy is converted to heat as the particles and gamma rays collide with the atoms that make up the reactor and its working fluid, usually water or occasionally heavy water.

Not to mention the fact that common nuclear fuels are very heavy elements, making them that much more efficient for weight/energy produced.

There was the proposed project Orion. Never got off the ground though, er..., not funded.

That's intriguing. Thanks for the link. ^^

Also, I stumbled across this that's a little more on-topic than all of our tangenting we've been doing lately. Very cool.

 

[Mark R]

My problem is with our focus on how we're converting the energy in such an energetic phenomenom as nuclear fission. Instead of utilizing the kinetic energy directly, such as that of a combustion engine or jet engine, we're simply harvesting the by-product that's produced: heat. Am I wrong? Is heat not the by-product? I have a difficult time understanding how such is the case if in chemical reactions the by-product is more-less the bane of the process rather than the product that we're aiming for.

You miss the point. Heat *is* kinetic energy - it is just the aggregate of a large number of particles with random kinetic energy. It can be converted to ordered kinetic energy (e.g. thrust from a jet, or movement of a piston) but at the cost of only being able to capture a fraction of the energy.

When a chemical reaction takes place, the various molecules gain kinetic energy as a result. However, as the molecules are randomly oriented at the time, the reaction energy is dispersed randomly (hence is heat). Same thing with a nuclear reaction, the nuclei fragments recoil, but as the nuclear fuel is dense, they collide with numerous electrons, scattering them widely in different directions, and the electrons lose their kinetic energy radnomly in innumerable collisions - giving heat.

In the case, of a jet engine or internal combustion engine, it is convenient becuause the reagents (fuel and oxygen) are mixed with air (mainly nitrogen), so the heat transfer from reaction to working fluid is rapid and simple (as they are already mixed).

In a nuclear reactor, the nuclear fuel doesn't mix with air or gas - so an additional step of heat transfer is required - from fuel to gas, or water. If you want to make a nuclear jet engine, you can, and it has been done, and it does work. You simply use a nuclear fuel to heat the air, instead of fuel.

 

[evilspoons]

I agree with you, heating an electric car by using the batteries would not be a good idea due to the high drain on the batteries. While it is certainly true you would need at least SOME heat to melt ice from the windshield, and to make the car somewhat comfortable to ride in, why does the heat for the car have to come from the batteries? Why can't you have a propane tank and a small heating unit to supply heat to the vehicle? You could even use the propane to heat the battery compartment. If the battery compartment were well insulated I doubt it would take much propane to heat it and keep it warm.

The Volvo C30 Electric has an ethanol-fueled heater.
http://www.engadget.com/2011/03/25/volvo-c30-electric-test-drive-video/

For what it's worth, many cars already have auxiliary heaters (i.e. not from engine heat)... just not so much in North America. I think a lot of older air-cooled Volkswagens burned gas as a heat source.