openCortex: ARTICLE: The 3DB and solar cars




		Sep 3 4:44pm PDT

HOME	Login \| New user? Register		Help
	Don't know what openCortex is? Learn about it.


		Home » Stanford at ASC 2001 finish » The 3DB and solar cars


	PARENT CONTENT Stanford at ASC 2001 ...

LINKED PROFILES Third Degree Burner '01	ARTICLE The 3DB and solar cars OCID cn00000160	LINKED CATEGORIES Aluminum Batteries Composite Monocoque D... Solar Cells Steel Titanium See all 6 related categories

	LINKED CONTENT


				Link to...

		+Article \| +Photo

The 3DB and solar cars
by eloy avila
created 5/13/2002
submitted 5/13/2002 01:39:59 PM

Here is a paper -briefly- covering topics of interest to beggining solar car enthusiasts. Some parts are outdated, but obvious in that sense.

Here we go. I will try not to bore you but rather fill you with exciting facts about the world of solar car building. By now you should be somewhat interested in solar cars and should hopefully know that they are composed of certain parts. Throughout the summer I dabbled in a variety of parts. These parts include the photovoltaic cells, batteries, composites, welding, alloys, the electrical system, some SolidWorks (a cad program), and some other miscellaneous stuff. I will cover a few ask me about the others and I may be of help.

PHOTOVOLTAICS —
A solar car gets energy from sunlight. Photovoltaics or solar cells, turn sunlight (averaging up to 1kW per square meter) into electricity by means of a pn junction. You have a wafer of semiconductor material such as silicon (Group IV, - four valence electrons) that is basically neutral as a pure element. When you add a few parts per million of an impurity such as gallium (Group III) you get a p type semiconductor. Add an element from group V such as arsenic and you get an n type semiconductor. You unbalance the crystal lattice by adding the dopants. One, the group III elements, leave the semiconductor wanting an extra electron and the other, the group V elements, leave the material with an extra electron. This leads to the p and n type materials. When you sandwich then together you get the pn junction and a potential develops. One has excess electrons and the other has vacant spaces needing electrons called holes. At the immediate junction neutrality develops but a potential exists since the two materials are constantly having holes or electrons freed due to room temperature or light. When light hits a surface of the semiconductor sandwich these electrons/holes want to unite and create neutrality again; what better way to let them unite than through a circuit where they can do some work. The more light the more carriers or electrons and thus electricity. These material however respond best to photons of some specific bandgap energy level to create free carriers. Any photon below is unused (transmitted) and most in excess serve to create heat in addition to electricity.

Germanium (Ge) has an energy gap (Wg) of 0.67 electron volts (eV) which is in the infrared region of the spectrum: silicon’s Wg is 1.10 eV, also in infrared: gallium arsenide (GaAs) has an Wg of 1.35 eV, near infrared and gallium phosphide (GaP) has a Wg of 2.25 eV in the visible part of the spectrum. Thus we can see why germanium and silicon are good choices for solar cells. They absorb a very large part of the spectrum. It appears that they would convert almost all available sunlight into electricity but there are other little facts about solar cells the prevent us from harnessing all energy available in light thought solar cells. These materials are most efficient within a narrow band of light around their energy gaps and then decrease even though the light may be of an energy level or frequency above the material’s Wg.

Overall the current efficiencies (η) of solar cells vary from about 10% to about 32% depending on material and implementation. The lowest efficiency cells are currently the thin film cells on the market although they are cool in other ways. They can be translucent and come in several colors. Most are made with amorphous α Si and η will drop within a few weeks due to the instability of this type of silicon. If you are going to get them make sure you get the stabilized efficiency not the initial one. Stable efficiencies will range from 8% to 11%. Check out the flexible cells by United Solar at www.ovonic.com/unitedsolar.html, a part of Ovonic, who also makes nice NiMH batteries. Other thin films are made using titanium dioxide and dyes. These also degrade but can be made in nice shades perfect for window and wall type applications, just ask the folks of Titania cells at STA Australia or Solaronix (www.solaronix.com) also look into the transparent cells made by Toshiba. Sunways Tubingen makes a cool clear Power-Cell that may be perfect for a bubble.

Next we have the good old silicon cells. These go from low teens to low twenties in efficiencies. They are made from either multi-crystalline or single crystal wafers.

Silicon responds to frequencies in the range of 45μm to 120μm with a peak at around 90μm. Single crystal cells are better since they have less crystal lattice interruptions and thus have better flow for the carriers. You can tell a cell is polycrystalline by the shiny and scattered changes in light reflection. Both cells are a cool indigo blue depending on how they are encapsulated and processed. Most have metal electrical contacts on the front and rear surfaces. Sunpower Corporation (sunpowercorp.com 408-991-0900 with η up to ~22.5%) makes “backside contact” cells. These have both electrical contacts embedded within the rear of the cell. This maximizes η due to increased surface area. To minimize reflection they along with others like Universtity of New South Whales (UNSW), Kyocera, Spectrolab, Siemens, Solarex, BP Solar and ASE use thin layer of titanium oxide on the top surface o the cells. Some have also implemented a textured surface to increase surface. The leaders in silicon cell technology are Sunpower Corp. in Sunnyvale California run by former Stanford electrical engineering professor, Dr. Richard Swanson (408-991-0908) and Dr. Pierre Verlinden (408-989-2524) and UNSW in Australia. They both currently make single crystal babies with average η of over 22% but for an eight square meter array that means roughly $460K at Sunpower.

Silicon cells degrade in efficiency with temperature. When the car is moving past a certain speed the cells are cool enough to be optimal. We will be doing tests on this stuff with Mr. ERic Hultgren shortly. We cool them with cool water on the go but the interface between the silicon and the encapsulant may be strained to the point of crystal breakage with sudden temperature changes. These questions will hopefully be answered soon. To get more in depth specs on any cells by just about any manufacturer check out the master index at the Photovoltaics Testing Lab (PTL) at www.nrel.gov/ncpv/ or the Arizona State University testing lab data at www.asu.edu/east/ptl/asuptllinks.html. There are also two excellent articles, “700 High-Efficiency Cells for a Dream” Progress in Photovoltaics and Applications, Vol. 2. pp 143-152 (1994) P.J Verlinden, R.M. Swanson and R.A. Crane. The other one is “Pilot Production of High-Efficiency PERL Silicon solar Cells for the World Solar Challenge Solar Car Race,” Technical Digest 9th international Photovoltaics Science and Engineering Conference, Miyazaki, Nov. 1999, pp. 65-66. J. Zhao, A. Wang, D.M. Roche, S.R. Wenham and M.A. Green.

Other than silicon there are many other semiconductors that serve as solar cells. Gallium arsenide and germanium cells are worth mentioning since they have high efficiencies and superior temperature responses as compared to silicon. Gallium arsenide is usually paired with gallium indium phosphide and can have efficiencies ranging from 23% to about 30%. Germanium also has the added benefit that η increases with temperature. Most values for η are given under standard 1.5 atmospheres and 1kW/m2 at 25˚ C. Germanium cells will go up slightly in η when the temperature increases. The main drawback is price and unavailability presently. These cells are mainly put into orbit and less used terrestrially. Learn more about these cells from folks at Kyocera, Spectrolab, TecStar (www.tecstar.com), Kopin, and Japan Energy. For more info check out a nice article at www.nrel.gov/ncpv/pdfs/large.pdf.

Finally we have combinations of several semiconductor junctions in an effort to maximize the conversion efficiency of the spectrum. Multi-junction cells have achieved production efficiencies of up to about 32% at Tecstar and Spectrolab with the potential to get up to about 50% with a four junction cell. This may sound to good to be true and for us solar car racers it is. These cells are made for space use where clouds and varying spectral irradiance is not a problem. These cells are internally connected in series and act like batteries. This means that when clouds cover the sky some of the spectrum is worthless and one or more junction in your super cells show it by converting very little or nothing at all. Since they are all in series you array ends up being only as good as the worst performing cell junction. If you had perfect skies all the time you would well but unless the cell junctions are separated and optimized you will have a very expensive array with a poor performance under clouds.

Now we get to optimization of solar cell performance. Solar cells are most efficient before they are cut out of the wafers in which they are made. This is because there are no physical edges where carrier recombination (carriers have lifetimes) can occur. Area is also maximized. When the cells are cut the perimeter relative to the active area goes up and η goes down. At edges carriers tend to recombine faster and not help in yielding useful current. There are a few ways of dealing with this problem that have yielded mixed results for teams in the past but are worth looking into. These include: making larger cells, shingling cells to cover the inactive borders (done by Honda in the past), changing the doping level, and having lower resistivity cells. The better cells seem to be about 120μm thick and have n and p type doping levels at 1.0 E+16 cm-3.

Certain steps can also be taking in assembling the cells into a module/array to optimize their performance. Most arrays will lose one to two percent of their cell efficiency when assembled. Using the correct soldering technique recommended by the manufacturer and not over-wiring helps. Careful protection through appropriate diodes is also a must. The encapsulation is also important. The Third Degree Burner currently has some polyester material perhaps EVA as an encapsulant. Others are Kapton, plastic substrate made by 3M, and Tefzel made by DuPont. One should make sure that these have a low reflectivity and that the boundary layer between it and the cell will not act like a mirror. The encapsulant should not block or absorb too many useful light frequencies and that sunlight will not affect it too much as some encapsulants will yellow with sunlight exposure. Mounting should maximize cooling and hopefully keep aero in mind.

BATTERIES —

Unless you have only a few thousand dollars allocated for an energy storage system do not consider lead acids any more. Nowadays teams are getting lighter and getting more kilowatt hours (kWh) of energy stored. This is because battery technology is approaching incredible figures of capacity and performance. I will focus on the major ones here as there are too many to list here. If you are enticed by some other battery out there that you think suitable for solar car racing purposes do not hesitate to look into it and tell others.

Lead acids, with a cell voltage of 1.9 volts, will typically have energy densities around 35Wh/kg and power densities of 412W/kg. They have a sloping discharge curve and fairly long charging times (~5-15 hours) with cycle life of about 500 to 600 cycles to 80% depth o discharge (DOD). Their self discharge rate is about 5% a month up to about 20%. They currently cost less than $100/kWh but can be dangerous because of lead and rapid charge/discharge yielding high temperatures and hydrogen gas.

Next in line we have the Nickel Cadmium (NiCad) cells. These stand at and just above 50Wh/kg and boast a potential cycle life of up to 2000 cycles. They will unfortunately self discharge up to 100% in one month. Their cost has dropped to no more than $300 per kWh and their charging time is usually about 8 hours. Their charge/discharge curves, at 1.25v, are fairly flat and suddenly drop when the cell is discharged. NiCads tend to have an undesirable “memory effect” that hinders their ability to charge/discharge to full potential if the cell is not properly charged/discharged regularly every time it is used.

Now for what may be the longest spiel on any one battery you will ever want to hear. Buckle up and pause to digest every now and then. The nickel metal hydride (NiMH) battery (currently in the 3DB) has been a decent technology suitable for electric vehicles. With an energy density between 70 and 80Wh/kg, power density of 220W/kg, and a cycle life of over 600 this battery had been the best money could buy. Although it was over two times the price of NiCad batteries NiMHs have dropped in price due to new emerging technologies. They have a cell voltage of 1.2 volts and discharge curves similar to NiCads as they use similar chemistries, except for the negative electrodes. Their curves are very slightly sloping from a full charge of about 1.4-1.5v to a full discharge at about 1 volt. Performance is best at a charge/discharge temperature of between 20 and 23˚ C but charging at lower temperatures may be better. According to D. Berndt’s Maintenance-Free Batteries, capacity goes down as temperature goes down. He gives the following formula:

Ca = actual capacity at 20˚ C
Ca=C/1+δ(Θ-20)
where
C = capacity measured at 20˚ C
Θ = T in ˚ C
δ = Temperature Coefficient in ˚ C
(The temperature coefficient for NiMH cells is ~.7mV/K.)

When discharging 20˚ C is optimal for any rate but the faster the rate the less capacity achieved. At a one hour discharge rate, written 1C(A), and quite fast for a solar car, you will have a maximum capacity of about 97% at 20˚ C and about 90% at 60˚ C. If you are going to discharge the battery faster than the .3CA rate, that is faster than 3.3 hours (the rate at which maximum capacity is possible), then capacity will plummet at temperatures below about 10˚ C. Only 40% capacity is available at –20˚ C when discharging at the 1CA (1 hour) rate. Capacity will drop about 2%/10˚ C from the maximum possible (at 20˚ C) for that rate given temperatures over 20˚ C. Clearly it would be wise to keep all cells in a module at a minimal temperature difference. No greater than 8˚ C range amongst cells is recommended. Self discharge is at up to 30%/month but less at lower temperatures as with most batteries. These cells can be recharged in less than about 6 hours if done properly. The recommended method is that of DT/dt-Peak Voltage detect and Timer. At higher charge rates of less than about four hours the cell voltage will peak to 1.5+volts and then drop. When the voltage peaks capacity has been reached. At .1C(A) or ten hour charge rates and slower, the cell voltage will not peak and drop but rather will continue to increase to capacities over 100%.

These batteries are relatively safe unless opened and will not be damaged if overcharged every now and then since the evolved hydrogen is readily absorbed internally. This does not mean one should pump current into them until they explode because metal hydrides can be very reactive with air and high temperatures with hydrogen do not help. Companies in the NiMH business include GM Ovonic, Varta, Gold Peak in Singapore, Saft in France, Sanyo and Panasonic.

Finally we come to the ultimate in battery bait for the solar car fanatics — LITHIUM. Although most of these cells can be explosive if mistreated they have the potential not only to help kick butt in races but also to give commercial EVs a nice push into competing with ICEVs. Currently lithium chemistry batteries range in energy densities of 100 to an alleged 300Wh/kg. There are now many slightly dissimilar technologies all involving the lithium electrode in one way or another. The most common one is the lithium ion cell used in laptops and other such small electronic devices.

Lithium ion cells yield about 110 to 130Wh/kg at 3.6v and a cycle life of up to 1200 charges/discharges. They have similar temperature operating ranges as NiMH (-20˚ C to +60˚ C) and their discharge curves are also similar. Charging is done with the constant voltage-constant current (CV/CC) method. Careful monitoring is required because although these cells can be recharged in less than three hours, if not properly charged they can overheat quickly and even explode. For this reason lithium ion cells will come with protection circuitry as part of the package. Careful care will allow these batteries to live up to their expected 1200 cycles. As an added benefit, these things will lose no more than 10% charge per month; most lose less and none of the lithium cells have memory effects. Check these babies out at companies like 3M, Samsung, ElectroFuel, Kokam, and Battery Engineering Inc.

Lithium polymer cells are lighter and more flexible than lithium ion cells. Li-poly cells are made from thin sheets of material comprising electrodes and electrolytes all sandwiched together into virtually any form factor. These cells deliver presently commercially 120 to 180Wh/kg but there are companies out there with much higher numbers that have tweaked the chemistries here and there to get ahead. The operating voltage is very similar to lithium ion’s, normally 4.2~3.0v with a nominal 3.7v. Operating voltages vary from manufacturer to manufacturer and application so be aware of that with these cells. These cells have no liquid electrolyte, so they are much less explosive than the lithium ion cells. They have similar charge/discharge characteristics but most can operate within a more extended temperature range than most other batteries. They can deliver quickly large currents, like NiCads. They will provide better power at a temperature range between 60 and 80˚ C as done by 3M with their EV module. This of course is another plus for the hopefully hot sunny racing days across the US or Australia.

Lithium polymer and its many family members have taken the world by storm. Many companies are joining the bandwagon and producing these. There are special ones like lithium sulfur that have shown energy densities of 220 to 200Wh/kg. Look into companies like Moletech, PolyPlus, Valence Technology, Toshiba America, Lithium Power Technologies, and PolyStor. Also learn hoe to deal with numbers relating to energy and such and familiarize yourself with the chemistry you use VERY WELL. Try The Independent Energy Guide by Kevin Jeffries and crunch some numbers for practice.

COMPOSITES —

Now, I realize that so far the excitement is almost unbearable but there’s more yet. A large part of any highly efficient solar electric vehicle will be composed of composite materials. The term composite will describe many very different materials. The one thing all composites have in common is of course that they are made of two or more different materials so as to form a much superior structural material. This rough description will hopefully allow you to understand why they are so applicable. They can be made to do or replace virtually anything, usually with added benefits. There are things like GRP (Glass Reinforced Plastic) and FRP (Fiber Reinforced Plastic) used in boat hull construction. There are metal matrix composites, plastic/metal composites used in new laptops and cell phones, and even wood laminate type composites. Two good books: Handbook of Composites, ed. Lubin G. and the other in case building a composite amphi vehicle seems plausible is: Design of Marine Structures in Composite Materials, by C.S Smith.

The 3DB has a composite “tub” and shell body. These are made primarily of epoxy, Kevlar, carbon fiber and honeycomb filling. We will begin with the carbon fiber. There are generally two types, the HS (High Strength) and HM (High Modulus) variety. Some favorable carbon fibers include HM P-75S or P-129 and the HS Carbon Fortafil F-5. As with most fibers you can get carbon in balanced or unidirectional form. Unidirectional is generally stronger but more expensive. Carbon/epoxy composites tend to have fiber volume fractions (Vf) of between 0.5 and 0.62 depending on the type of carbon being used and the structural qualities desired. Specific gravity (SG) for carbon/epoxy laminates runs from 1.5 to 1.7 and achieve Young’ s Modulus E ranging from 55 GPa for the HS balanced 0.5 Vf laminate to 300 GPa for HM unidirectional 0.62 Vf. For comparison, marine mahogany plywood has an SG of 0.6 and steel runs at around 7.8. Their Young’s Modulus are 7 GPa for the plywood and around 207 GPa for steel. Combinations of carbon composites can be made tailored to have high shear modulus, tensile strength, compressive strength, or shear strength or more than one.

Carbon, specifically HS/epoxy at Vf =0.5 will have a very low coefficient of linear expansion (1x10-6/˚ C parallel to fiber direction) as compared to 5083 aluminum’s 22 x10-6/˚ C. Thermal conductivity of this same combination is at 0.7 W/m ˚ C perpendicular and 10 W/m ˚ C parallel. Carbon fiber is also slightly conductive so don’t make battery boxes out of it. Because it is so stiff it tends to do very poorly during vibration. Carbon however tends to retain most strength and stiffness at temperatures (relative to 20˚ C) to 1000˚ C provided oxygen is excluded and only at temperatures around -200˚ C will it lose 10 to 25%. Whereas fiberglass or E-glass will reduce in strength and stiffness by about 75% at 350˚ C but will increase at lower temperatures. Carbon, by the way, also performs well in aqueous or acidic environments.

Another material used in the 3DB is Aramid Kevlar 49. This is the yellow ochre fabric used in bullet proof vests. Unimpregnated (without resin) fiber of this kind 26mm thick is enough to stop an armalite bullet at 55g with a velocity of 976 m/s. This fiber is paired with carbon to compensate for bending and compression and also as a high vibration damping factor critical in top shell construction where brittle solar cells reside.

Relative to 20˚ C Kevlar 49 will retain 75% strength and stiffness at 200˚ C and 60% at 300˚ C. At lower temperatures this fiber will retain these qualities quite well. It will perform less well than carbon in water or acid but can be paired with CEM-FIL fiber or E-CR glass for this purpose.

A Kevlar49/epoxy unidirectional laminate at 0.62 Vf will have a specific gravity of 1.4 with a Young’s modulus at 50 GPa, shear at 8 GPa and an extremely high tensile strength of 1600 MPa compared to carbon’s 360-1500 and aluminum’s (5083) 150 MPa. A laminate of 0.48 Vf will have a zero coefficient of thermal expansion and thermal conductivity at 0.22 W/m ˚ C perpendicular (similar to fiberglass/polyester) and 0.9 W/m ˚ C parallel. It also has a high volume resistivity at 5x1013Ω/m (compared to even E-glass/polyester at 2x1013Ω/m). Typically dry weave Kevlar K-120 plain at 1.8oz/yd2 and 5 mil thick or K-285 proflux satin at 5oz/yd2 and 10 mil (used in 3DB) are used for the shell. You can get these dry or unimpregnated or prepreg with a certain amount of added resin. Prepreg will come frozen to prevent premature curing and has a lifetime. Once thawed a small window within which it must be worked exists. This method is generally best for more uniform results and less hassle. If you are a composite guru then working with dry weaves and the right amount of the right resin you can achieve much better results. If you are not too sure about what you are doing however, you may be lucky to have something that will not fall apart very soon after it’s cured. There are several ways to go about curing composite parts with Kevlar. I have learned that the shell of the 3Db was made with prepreg cured at 250˚ C or 200˚ C with a vacuum bag. It is best to use a vacuum as this will help to get a more uniform bond and if there is excess resin it will be expelled. Higher temperatures will cure faster and a high vacuum will also do the same but manufacturers will recommend slightly different techniques, look into them and ask the experts. A good way to judge composites is to compare toughness coefficients found by squaring strength and dividing by two times the modulus. (strength2/2 x modulus)

If you look at the shell and “cockpit” on the 3Db you will notice that between layers of carbon fiber or Kevlar is a layer of little honeycomb-like structures. If it is an brown-orange color it is Nomex honeycomb made with Aramid paper. It can be made with glass fabric (style 7781), Kevlar or Aramid (style 285) or graphite (style 30K70PW or T300) and in varying thickness with several wall orientations. The purpose of these honeycomb fillers is to add compressive strength while greatly reducing weight. These sandwich core materials will typically have specific gravities (SG) of as low as 0.065 for Nomex Aramid paper to no more than 0.19 for PVC or PU foam core. As far as structural numbers go, the best performers for core materials, in order from best to worst, are aluminum honeycomb (with an SG of 0.07), Nomex Aramid paper, End-grain balsa, and PVC foam with an SG of 0.19.

Now for what little I know about the stuff used to bond composite constituents.

Epoxy usually cements these materials together to form the composite. These are typically thermosetting. Epoxy has an SG of 1.2 and a Young’s Modulus of 3 GPa and tensile strength at 85 MPa and compressive at 130 MPa. When cured it will distort at a temperature of 110˚ C. Epoxy is favored above polyester resins because it does not blister as much. Polyester resins are also attacked by chemicals like acetone and trichloroethylene. Like epoxy and vinyl ester, polyester resins will burn slowly in air with a lot of black smoke but can be put out with water. Phenolic resins have a higher ignition temperature and will burn with lower smoke evolution if you are worried about fire hazards. For more info on these check out page 26 of the Design of Marine Structures in Composite Materials Book by C.S. Smith.

METALS AND WELDING —

Many parts of the 3DB and any solar racer will have to be made of metal. Whether it is the grade five or grade eight bolts, the roll bar or the suspension among many others. I will focus on a few major ones. I currently know nothing about machining them, ask Mr. Joel Segre or Mr. Champ about that.

ALUMINUM is used all over as the suspension and many structural brackets. Aluminum stock weighs about 166 to 177 pounds per cubic foot depending on the alloy. Typical alloys used are:
2024 T4: easy to machine but impossible to weld and a shear strength of 41 ksi.
6061 T6: very easy to machine, and able to weld and shear strength of 30 ksi and weighs 170 lbs/ft3. (currently this is what the hubs of the 3DB are made of.)
7075 T6: fairly easy to machine, hard to weld and shear strength of 48 ksi. It weighs in at 174 lbs/ft3. (currently this is what the suspension A-arms of the 3DB are made of.)
7175 is similar to 7075 but a bit stronger.
7178 T6: fair to machine, very hard to weld and shear strength of 52 ksi and weighs 177 lbs/ft3.

STEEL is not used where possible because it weighs more than aluminum but if needed it would be wise to use stuff like Tool Steels Vasco Jet 100 or 882 or H11 used in airplane landing gear. I have welded several steels in the 1000 series. This series is pure steel with certain percentages of carbon. 1080 for example contains 0.80% carbon and 1018 contains 0.18% carbon. Welding anything above 1/32 of an inch thick with the wire fed MIG at the site is easy. Follow the instructions under the lid and adjust accordingly after some practice to reduce the crackling sound you will get so it sounds like a more smooth hiss. First, make sure the two pieces fit well and are clean. Keep the wire feed so that you are moving fast enough to not cut through but securely welding the two pieces. There should be no more than about a 1cm length of wire between the tip and the weld pool at all times. It is best to keep the tip perpendicular to the weld unless you are at an angle or vertically then adjust to prevent dripping and cutting. The weld should end up smooth and even, slightly rising above the rest of the surface. There will be a brown and or white residue from the flux around the weld. It should extend no more than a few centimeters beyond the weld so you know you are getting a good arc/melt/weld. Don’t be afraid to go over suspicious welds. Better safe than sorry. Some metals will be weaker at the joint when welded and must be heat treated afterwards if total strength is to be regained. This is the case with some steel and all aluminums.

Make sure that the wire will also not need heat treating after the weld since it may be of a different alloy. The wire should, of course, match the alloy you are welding. Aluminum is welded with a MIG (Metal Inert Gas) or a TIG (Tungsten Inert Gas) welder. The inert gas is pumped out through the tip and is either CO2 or argon or mixtures of these with oxygen. These help the metals melt without impurities because the gases shield the tip as it welds provided you don’t have a foot of wire between the tip and weld pool. A TIG is used to weld aluminum and titanium alloys. It is a tungsten tip that melts the two parts together without a wire. If a filler is needed, a similar alloy is added at the same time as in brazing with an oxy-acetylene torch (good for 1/16 and thinner stock).

Proper welding procedure is a must—don’t blind anyone, burn anyone, electrocute anyone, and remember to have someone with you while you weld. Wear proper clothing as hot metal burns and high intensity UV will yield quick and deep tans.

TITANIUM is not currently used in the 3DB but I think it should because it is cool albeit very expensive. Whereas steel weighs around 490 lbs/ft3 titanium weighs about 281 lbs/ft3. It could be used for the roll bar to replace the steel one currently used and not properly attached. Good alloys include:
5-2.5 used in aircraft engines and airframes. It can be machine and welded.
6-6-2 used in airframe sections and rockets and ok to machine and fair to weld.
8-1-1 is used in jet engines and has the highest tensile strength and lowest density of titanium alloys. It is also fairly easy to machine and weld.
Ti 15-3 is a high strength alloy used in aerospace structures that is fair for machining and good for welding.

Let’s compare the roll bar with several alloy tubes of equal dimensions equal to an open diameter of 1 inch 1/8in thick and 60 inches long. Pure grade titanium would weigh 5.76 lbs, Aluminum 2024 would be 2.06 lbs, 6061 would be 2.02 lbs and 7075 would weigh 2.084 lbs. One must remember that a certain strength would be needed for the roll bar and to achieve this more aluminum would be needed. Steel would be about nine or ten pounds so going with titanium would save you weight that can add up if the material usage is optimized. Bolts and washers may be replaced with aluminum ones without considerable strength loss where possible.

By now you are either a bit more enlightened or almost asleep. There is more to this than I have written here for the sake of getting this done and poor writing in some of my notes. Have questions and have the gumption to ask them or look for answers. The car will not build itself and the more we learn about what goes into it the more we will get out of this project. Get interested in something, small as it may be. I hope this helps get you started or refreshed.

5633 words | jasari

How informative is this Article?

Quality Rating
Overall: Very Informative
Rated by 3 members:
aselkirk Very Informative
nkosta Very Informative
solarpaul86 Very Informative



	Related Content (1)

	Stanford at ASC 2001 finish Photo by eloy avila created 7/25/2001 submitted 5/13/2002 01:14:31 PM Here is the entire Stanford #16 team at the ASC 2001 finish in Claremont.