Loading summary
Transcript1 lines
- [00:02]
A
A gas turbine's high pressure rotor takes on some of the most extreme temperatures in industry and must do it for 100,000 operating hours. Right now, we are booked out of gas turbines for the rest of the decade. And the shortage got me thinking about fan blades. Today, modern gas turbine inlet temperatures can reach an infernal 1600 degrees Celsius, which is hotter than most lavas. What can handle such extreme conditions? Some of the most special metals you will ever see in your life. Metals tortured to survive. Things that regular metals never can. And it's still not enough. In today's video, my friends, we study the blade and the materials that make them. Gas turbines and jet turbines too. Take in air current, compress it, and then add fuel, usually natural gas. A combustor then ignites the combination to bring it to a high energy state. That high energy air hits the turbine blades, transferring their energy to the blades and motivating them to turn. The hotter the gas is before it hits the fan blades, the more efficient the turbine will be. The more efficient the turbine, the more fuel savings. Depending on the size of the plant, one efficiency point yields up to $25 million in lifetime fuel savings. But higher inlet temperatures can only be achieved if we have blade materials that can tolerate them. Particularly for those blades positioned right after the combustor, these puppies experience the hottest temperatures and highest pressures. There are some other concerns. Turbine blades are exposed to a fair amount of corrosive gas and centrifugal forces. Blades do about 10,000 to 12,000 rpm, which exerts an outwards pull of about 20 tons per square inch, or 276 MPa. The main long term concern with that is something called creep. Creep refers to deformation that occurs over extended periods of time to to the metal being exposed to long periods of centrifugal stress. At high temperatures, eventually the blade might creep so much that it scrapes the casing or just outright fractures. Regular steels are not suited for such high temperatures. Stainless steel's melting point is between 1370 and 1540 degrees Celsius, depending on the alloy. Sounds high enough. But steels lose their strength far before they hit those temperatures. At just 50 to 60% of its melting point. So about 540 degrees Celsius, the steel rapidly starts turning into something like taffy. You get insane creep. So we need something special. For such applications, the industry turns to metals that we call superalloys. There is no hard guideline as to what qualifies as a superalloy. Some have suggested it to mean alloys that retain long time strength at high temperatures, about 70% of their melting point, while also being resistant to creep and oxidation. I reckon that works. The name was coined in the 1940s, Superman and all. But the concept dates to the early 1900s. Back then, an American metallurgist named Albert Marsh was searching for metals with high resistance that can be used to make heating elements. He discovered an 8020 nickel chrome alloy that fit the bill and patented it in 1906. They dubbed it Chromel, but we call it nichrome. Marsha's cheap nichrome wire based heating elements help make the toaster possible. That's a legacy I would want for myself. Nichrome can be considered the ancestor of today's nickel based superalloys. There are over 100 today. In addition, there are two other Rough family categorizations. Those superalloys based on cobalt and iron Nickel chromium. Cobalt is rather rare, so most commonly used superalloys are nickel based. Nickel is relatively plentiful thanks to mines in Australia. Its melting point of about 1,455 degrees Celsius is very high. Better yet, it retains strength at up to 85% of that melting temperature. In the 1940s, people started working on producing the first jet engines because of, you know, the war thing. Meanwhile, metallurgists were making strides in understanding how different metal structures lead to different properties. In 1920, researchers realized a relevance of creep for things like steam boilers and began studying its causes. By the 1930s, we recognize the importance of crystal structures in determining a metal's physical properties and creep resistance at high temperatures. Now I warn you, this is complicated, so I'm going to give a simplified version. I pray this is right. Turbine blades are casted. We melt down the superalloy metal, pour that hot melt into a mold, usually a ceramic one, and then cool it down. This conventional casting method gives you a metal made up of many small crystals called grains. I will use the two words interchangeably since the crystals tend to be about the same size. We call them equiaxed grain. The interfaces where different grains meet, where the crystal lattice orientations do not match up and change direction are are called grain boundaries. Creep can happen at two levels. Between the crystals, at the grain boundaries or within the crystal grains themselves. If we want to harden the superalloy's creep resistance, then we must address from both approaches. Imagine the crystal grain itself has been made up of planes or sheets of metal atoms stacked together. This is a crystal lattice. Forces pushing on the Crystal lattice will create dislocations, meaning defects or wrinkles in that crystal lattice. Over time, the dislocations start moving throughout the lattice. And that is how we get the one form of creep. So preventing this means finding some way to impede these dislocations as they travel through these planes of metal atoms and in some way adding some form of friction. So we turn to a concept called precipitation hardening. This is a strengthening technique that can be used for all metals, including nickel based superalloys. The idea behind precipitation hardening is to add special precipitates called gamma prime into the alloy grain. These gamma primes are quite special. Grains are made up of unit cells featuring a five atom cubic structure. Four atoms at each corner with one in the center. Normally you can have either an aluminium or nickel atom mixed up at any one of these five positions. These cubical arrangements or phases are called just gamma, not gamma prime. Gamma prime on the other hand, is a very specific phase. Four aluminium atoms at the corners and one nickel atom at the center. This is way more ordered. Gamma prime adds friction between the sheets of metal atoms, thus making it harder for dislocation to travel through it. Impede the dislocations and we hinder and slow the progress of the creep. We perform precipitation hardening in a three step process. First, we add small amounts of metal additives to the melt. In the case of the nickel based superalloys, the additives are often aluminium and titanium. These are added at high temperatures so the additives can mix together in a single uniform solution, like sugar cubes in a cup of hot tea. After this, we quench, meaning to rapidly cool down the mixture. This is so that we can trap the additive elements throughout the nickel melt before they can diffuse out. Finally, we need to actually create the gamma prime precipitates. This is done by reheating the metal to a moderate temperature so that the aluminium, titanium and nickel can form the gas. Gamma prime precipitates we so desire. The first superalloy to be strengthened this way was Mnemonic 80. First produced by the Mond Nickel Company in 1940. This iconic superalloy remains in use today. The technical breakthrough that really unlocked the superalloy game, however, was vacuum melting. Prior to the 1950s, we melted down and created alloys in open air. Air is great. I breathe it all the time. However, air also contains highly reactive elements like oxygen or carbon. Making superalloys, we want additives like aluminium and titanium to react with the nickel. But when melted down in open air, the metal additives would instead react with oxygen or carbon, creating inferior metals like as if in a bad soap drama when the protagonist runs off with the sexy butler rather than enter an arranged marriage. Then in the late 1940s, a team at General Electric led by James Nisbett was experimenting with new alloys. For more consistent results, he started melting down the metals in a vacuum. At the time it was just good science. But it turned out to be the key step to better superalloys. Nisbet would later go on to found the company Allvac, now a subsidiary of Allegheny Technologies, or ATI, not to be confused with the GPU company. In 1950, Dr. Gunther Moeling of the Allegheny Ludlum Steel Company was the first to do vacuum melting at commercial scale. By the mid-1950s, the industry had accepted vacuum metallurgy as a critical method to producing good superalloys. Vacuum melting also paves the way for the next two superalloy breakthroughs, Directional solidification and single crystal. Note that I said that we can get creep either within the crystals grains or between them. In the latter case, creep occurs along the interfaces between the crystals, the grain boundaries. If we can somehow remove those grain boundaries, then we can raise creep resistance. This was achieved in the 1960s and 1970s with alloy processing rather than alloy chemistry. Meaning rather than trying to find new elements to mix in, they introduced new methods to post process an existing metal. Pratt and Whitney led the way by applying directional solidification. Interestingly, I remember talking about this method in two videos on very different subjects. Polycrystalline silicon for solar cells and scintillator crystals for PET scanners. Directional solidification involves pouring hot melt into a ceramic cast inside a high temperature furnace whilst in a vacuum. At the bottom of the cast is a water cooled chill plate sitting outside the surrounding furnace. The melt pouring in at the bottom touches the chill plate and and crystallizes. Then slowly and deliberately we move the cast and the melt inside of it downwards out of the furnace. If done right, then the melt crystallizes from the bottom to the top, growing these tall straight vertical crystal columns. The grain boundaries going left to right are largely removed leaving only those parallel to the blade's length running along the turbine blade's major stress axis. Pratt and Whitney first used directionally solidified superalloys to produce the turbine fan blades for the J58 engine in the Lockheed SR71 Blackbird plane. I am told that this airplane is mildly famous. After this, the obvious next step is to eliminate the grain boundaries entirely and make the whole blade a single crystal. The techniques to produce. These were pioneered at the same Pratt and Whitney research lab team that commercialized directional solidification, led by a brilliant scientist named Frank Versniter. While experimenting with the casting molds, they noticed that adding a right angle bend to the starter chamber mold above the bottom chill plate filtered out some of the vertically growing columnar crystals by blocking their growth paths. They discovered that adding more bends led to fewer crystals emerging over those bends. The team later swapped out the right angles with a smoothly curving helix shape. In industry, this helix is referred to as the pigtail. Anyway, the pigtail takes in the growing columnar crystals and through one or two turns, filters out all but the fastest growing single crystal. Then, as with regular directional solidification, we slowly move the melt downwards. That winning crystal exits the pigtail and eventually expands to crystallize throughout all the melt in the whole mold. The result, if done right, is a turbine blade arranged into just one single crystal. No grain boundaries though. I want to note that though the atomic lattice is unbroken, the crystal itself is not homogeneous. You still need those phases, those gamma and gamma primes, to add strength, which explains why they might still look like they have grain boundaries. Anyway, single crystal processing was a game changer. People went back to the drawing board and created new superalloy mixes specifically targeted to the method. Most would incorporate a very rare element called rhenium. Unlike the so called rare earths, rhenium is actually rare, the byproduct of a byproduct. They only produce about 50 tons of it each year. Adding rhenium to nickel based superalloys in proportions of up to 6%, dramatically improves creep resistance and overall temperature resistance for single crystal superalloys. Thus, the majority of the world's rhenium goes into such superalloys. In the 1950s and 1960s, superalloy research focused on jet turbines for military and civilian uses. Superalloys, among other technical innovations, helped triple those planes thrust to weight ratios, while also cutting down the time between overhauls from 100 hours to 100,000 hours. Though I'd say that military and commercial planes do demand different things from their turbines. Then in the 1970s, the oil crisis accelerated the advancement of gas turbines. Bagasse turbines require much larger blades, anywhere from 45 to 80 centimeters long. Trying to cast such massive blades exacerbated a persistent defect condition called freckling. Freckles refer to tiny equiaxed grains that form inside the single crystal. Since they are random grains with grain boundaries, they hurt the blade's strength. They happen because the alloy doesn't uniformly freeze during directional solidification, creating convection currents that eventually nucleate unwanted crystals. This required making entirely new furnaces with very tight control over the thermal gradients. Even with all this superalloy stuff, it is not enough. One of the most advanced combined cycle gas turbines today is is the Mitsubishi J series. With an efficiency of about 64%. Its operating inlet temperature is 1600 degrees Celsius. That is hotter than typical lava and three times hotter than a pizza oven. It exceeds the melting points of steel, cobalt and nickel, meaning that even the superalloys will melt at such temperatures. So how does the turbine not melt down? Engineers have incorporated cooling measures to maintain the metal's internal temperatures at about 80 to 90% of their melting points. One thing they do are thermal barrier coatings to reduce how much heat eventually reach the superalloy surface. This barrier coating usually ranges about 100 to 400 micrometers thick. These coatings can be just as complicated as the metals they protect. They have to remain on surfaces for tens of thousands of hours while enduring the metals underneath them, repeatedly stretching and shrinking with rising and falling temperatures. They're actually made up of multiple layers. At the top we have a ceramic topcoat layer with very low thermal conductivity. This serves as the primary insulation layer. Below that is a bond coat that adheres the topcoat to the superalloy surface below. Its number one job during high temperature operations. The bond coat will help grow a third layer in between the top and bond coats. An aluminium based thermally grown oxide layer to stabilize the adhesion as well as protect against oxidation. All of these are made from chemical compounds with enough letters to win a Scrabble game. Another measure is to use air to create a protective cooling film. Air cooling. These turbine blades are injected with cooler air routed through from the compressor. Cooler is only relative. At about 600 to 650 degrees Celsius, the air is quite hot, but it still beats 1600 degrees anyway. The air is injected into small passageways inside the blade and and exit through tiny laser drilled holes carefully arranged so to create a protective envelope of cooling air on the blade surface. It works, but designers should keep in mind that the gas turbine will need to exert extra energy to provide all the cooling air for the blades. Remember, the cooling air is constantly bleeding out. These are inefficiencies that cause the turbine to underperform its ideal. An interesting alternative would be routing airflow inside the blade. Using complex computer modeling, we can craft channels that produce rotating swirls of air to exchange heat with the superalloy's inner surface, keeping it cool. It's still inefficient, but maybe less so because we don't lose that air until it exits the blade anyway. Between these and the thermal coatings, the blade temperature is maintained at about 1150 degrees Celsi. I knew coming into this that these blade materials are insane, but gosh. Each turbine blade is made from superalloys of nickel plus up to 12 other elements, including one of the rarest elements on earth. Factories then must cast these into ultra complicated 3D shapes with internal passageways and circulating air nozzles using the lost wax casting technique. At the same time, we employ directional solidification plus pigtail to produce the blade as a single crystal. Then the single crystal blade must get its cooling air holes drilled. It must be heat treated and then it has to get its multi layer thermal barrier coating applied using E beam plasma deposition. Then we stick it on a rotor and expose it to the fires of hell. Alright everyone, that's it for tonight. Thanks for watching. Subscribe to the channel, Sign up for the Patreon and and I'll see you guys next time.