A history of the Van de Graaff generator

IEE

E I

ND

US

TRY

AP

PLI

CA

TIO

NS

MA

GA

ZIN

E •

JA

N|

FE

B 2

00

5 •

WW

W.I

EE

E.O

RG

/IA

S

10

h is tory | f.a. (tony) furfari

his month I writeabout an “atomsmasher,” the Van deGraaff generator. TheVan de Graaff genera-

tor is defined in my dictionary as “anelectrostatic generator in which elec-tric charge is …transferred to a largehollow spherical electrode by a rapid-ly moving belt, producing potentialsover a million volts, and used withan acceleration tube as an electron orion accelerator” [1]. It was inventedby Robert Jemeson Van de Graaff in1930 while a post-doctoral physicistat the Palmer Physics Laboratory atPrinceton University.

Robert Jemeson Van der Graaffwas born on 20 December 1901 inTuscaloosa, Alabama. His motherwas Minnie Cherokee Hargrove andhis father was Adrian Sebastian Vande Graaff. Robert attended theTuscaloosa public schools and thenattended the University of Alabamawhere he received a B.S. degree in1922 and an M.S. degree in 1923.Both degrees were in mechanicalengineering. After graduation, heworked for the Alabama PowerCompany for a year as a researchassistant. He studied at the Sorbonnein Paris in 1924–1925 and, whilethere, attended lectures by MarieCurie on radiation. In 1925, he wentto Oxford University in England as aRhodes Scholar. At Oxford, hereceived a B.S. in physics in 1926and a Ph.D. in physics in 1928 [2].

While at Oxford, Van de Graaffbecame aware of the deep interest ofnuclear experimenters in the possibil-ity that particles could someday beaccelerated to speeds sufficient to dis-integrate nuclei. By disintegratingatomic nuclei, much could be learned

about the nature of individual atoms.Van de Graaff focused on this facet oftechnology after attending a lectureby Sir Ernest Rutherford(1871–1937), at the Royal Society.During the lecture, Rutherford, dis-coverer of the atom, stressed thatthere was a tremendous need foraccelerating particles of controlledenergy in the study of nuclear physicsand implored the audience to doresearch in that field. Van de Graaff,

being a young scholar, took Ruther-ford’s admonishment very seriously,and was motivated to try to develop aworking system to achieve particleacceleration.

Van de Graaff Produces His First Machines For a discussion of how the VanGraaff generator came about andhow it works, I sought the help ofDr. Chatham Cooke at Massachu-setts Institute of Technology (MIT),and I share his descriptions with ourreaders. Dr. Cooke is senior lecturer

and principal researcher at MIT’sDepartment of Computer Scienceand Electrical Engineering.

In 1929, Van de Graaff, upon hisreturn from Oxford to the UnitedStates, joined the Palmer Laboratoryat Princeton University as a NationalResearch Fellow. During the 1930s,the movement of electromagneticcharges was recognized as a possibleprocess for creating very high volt-age. Systems that move charged par-ticles were envisioned by nuclearphysicists as possibly being liquid,and not necessarily solid, such as abelt. Corona spray had been recog-nized as a charge that could beapplied to a belt; but the questionwas how to collect the charge from amoving belt. Van de Graaff, being aphysicist, recognized that if you ran abelt between pulleys, one pulley atground and one at isolated environ-ment, supported by insulators, volt-age could be generated, but thedifficulty was to move charges off ofthe belt that is carrying the charge.For instance, if charging with posi-tive charges, when trying to bringmore charges up, the sphere is tryingto push them away. What Van deGraaff discovered was the concept ofusing a re-entrance structure, wherethe belt enters inside the sphere.That is the very key component,because once inside the sphere, thecharges are naturally drawn to thesphere, trying to get to the outsidesurface of the sphere. There’s a forceto move charges from anywhereinside the sphere to the external sur-face of the sphere. By bringing thepulley inside the sphere, which is thecharge-collector, it was very efficientand effective to remove the chargesoff the belt.

A History of the Van de Graaff Generator

Robert Jemison Van de Graaff.

T

CO

UR

TE

SY

MIT

MU

SE

UM

Authorized licensed use limited to: IEEE Xplore. Downloaded on January 18,2022 at 04:08:03 UTC from IEEE Xplore. Restrictions apply.

11

IEE

E IN

DU

STR

Y A

PP

LICA

TION

S M

AG

AZIN

E •

JAN

|F

EB

20

05

• W

WW

.IEE

E.O

RG

/IAS

Van de Graaff constructed afirst working model of an electro-static accelerator; it developed89,000 V. Improvements to hisoriginal design followed, and inNovember 1933, at a meeting ofthe American Institute of Physicshe demonstrated his machine thatproduced 1 million volts and wasonly a few feet tall.

Robert Van de Graaff Returns to MITKarl Compton, then president ofMIT, became aware of Van de Graaff’saccomplishments at Princeton andinvited him to return to MIT as aresearch associate. It was there thatVan de Graaff built his first largemachine in an aircraft hanger. Themachine used two polished alu-minum spheres, each 15 ft in diame-ter, mounted on 25-ft highinsulating columns that were 6 ft indiameter. The columns were mount-ed on railroad trucks that put thecolumns 43 ft above ground level.Each sphere had its own generator,one arranged to charge positive andthe other negative, so that the volt-age between the two spheres wastwice that of the generator alone. Themachine was displayed on 28November 1933 and produced 7 mil-lion volts. The accomplishment wasreported in a story in The New YorkTimes for 29 November 1933 titled:“Man Hurls Bolt of 7,000,000 Volts.’In 1937, the machine was placed in apressurized enclosure at MIT. Therewere two terminals in twomachines: one designed to run posi-tive and one designed to run nega-tive, and the acceleration tube wasset up to run between the two. Theintent was to have each machinerun at a couple of million volts and,therefore, get multimillion voltenergy that could facilitate radia-tion of materials.

In 1935, Van de Graaff receivedU.S. Patent No. 1,991,236 for hisinvention. In 1936, Van de Graaffmarried Catherine Boyden. They hadtwo sons, John and William. RobertJ. Van de Graaff died on the morn-ing of 16 January 1967 in Boston atthe age of 65. At the time of hisdeath, there were over 500 Van deGraaff particle accelerators in use inmore than 30 countries.

Compressed Gases Instead of Air Is BestMuch was learned about how to makemachines at that time, but it was rec-ognized that air was not going to be asuccessful medium. There was tremen-dous effort to develop gases that wouldbe better insulators. Carbon-tetra-chlo-ride was one of the early good ones,but it was unhealthy for humans,Unfortunately, besides beingunhealthy, it also has a low vapor pres-sure, so its molecule cannot be in agaseous state at high pressures, so itwas difficult to get sufficient density ofgas to make it electrically effective.

However, other gases and gasmixtures were explored, and high-pressure gases were used immedi-ately, nitrogen, air, and carbondioxide being the dominant ones.There continued during the early1930s a major search to develop abetter gas. Ultimately it was rec-ognized that sulfur hexaflouridewas a prime target for insulation,and researchers set about ratherquickly trying to find a way togenerate a high-voltage gas. Afterrealizing it was a good electricalgas, the first application, was inVan de Graaff’s machines.

It has been told that Van deGraaff spent about US$400 (whichwas a lot of money at that time) toget a few grams of SF6. The chem-istry department at MIT specifi-cally isolated it for Robert Van deGraaff, in order for him to make

tests for its use as an insultingmedia for his generator. He discov-ered it was a great gas for insulat-ing purposes. It was also not toxic,and it could be pressurized to acouple hundred psi, making itdense a terrific insulator. Also, instudying those properties, Van deGraaff recognized that, besidesbeing a good insulator, when arcsdid occur, they were quicklyquenched. Going into compressedgas environment quickly becamethe recognized way to get muchhigher voltage machines. Severalwere built at MIT and other

The engineer’s prototype of the generator being demonstrated by Dr. Van deGraaff at the Hotel Statler. It stood approximately 6 ft tall and was capable ofproducing about 1 million volts of static electricity.

This is an early Van de Graff generatorbeing demonstrated by Robert J. Vande Graaff himself.

CO

UR

TE

SY

MIT

MU

SE

UM

CO

UR

TE

SY

MIT

MU

SE

UM


IEE

E I

ND

US

TRY

AP

PLI

CA

TIO

NS

MA

GA

ZIN

E •

JA

N|

FE

B 2

00

5 •

WW

W.I

EE

E.O

RG

/IA

S

12

institutions. Additionally, becauseof its thermodynamic and arc-quenching properties, SF6 wassoon applied to high-voltage cir-cuit breakers.

Other Contributors to TechnologyVan de Graaff worked with John G.Trump, a professor of electrical engi-neering at MIT, and with WilliamW. Buechner, a professor of physics atMIT, developing a machine to use asa medical unit to produce X rays fortreating cancerous tumors with pre-cise penetrating radiation, first usedclinically in 1937 at Harvard Med-ical School.

The Carnegie Institution inWashington, DC, sometime around1935, developed a Van de Graaffaccelerator for about 1 million volts.The machine was developed byPhysicist Merle Tuve; it was airinsulated and proved to be quite asuccessful machine. At the Univer-sity of Wisconsin, physicist Ray-

mond Herb (1908–1996) produceda compressed gas insulated machineof the Van de Graaff type; it wasquite successful. Robert Van deGraaff and Ray Herb ended uptransferring their technologies ofelectrostatic machines to applica-tions in industry. The Van de Graafftechnology was implemented by acompany called High VoltageEngineering Corp. of BurlingtonMassachusetts. Ray Herb’s technol-ogy was implemented by a compa-ny called National Electrostatics,just outside of Madison, Wiscon-sin. Both companies producedmachines for many decades andwere responsible for providing thetechnology that served nuclearphysics for many years.

In 1948, an article summariz-ing some of this history was pub-lished by the American PhysicalSociety, titled “Reports onProgress in Physics,” authored byRobert J. Van de Graaff, John GTrump, and William Buekner, all

of MIT. Trump was an electricalengineer and a close friend of Vande Graaff. Van de Graaff was aphysicist, most interested in hav-ing the machine for physicsresearch; Trump, on the otherhand, was an electrical engineer,and saw a broader application forthe machine. Trump turned todeveloping the machine for radia-tion therapy and became a leaderof applying the machines to hospi-tal environment, where cancertreatment with X rays had beenperformed for decades.

Applications of the Van de Graaff Machine The machines at MIT were used forphysics research, a first targetbecause of Van de Graaff’s devotionto nuclear physics. After that,researchers recognized there werepossible commercial values, causingchanges associated with radiation,and using electrons to generate Xrays by having them strike a heavytarget. Some machines were appliedto X-ray imaging, several machineswere used by the government dur-ing WWII to X-ray interiors ofnavy ordinance. The high-energy Xrays have tremendous power and canpenetrate very deeply. Also, by cre-ating a very intense, highly focusedbeam, X rays can be created thatcannot be created from a radioactivesource such as cobalt. Furthermore,because they can be pinpointed tocreate very high-resolution X rays,the source is a sharp point. X-raytubes running up to about 250 kVwere common, but getting higherenergy tubes was not possible at thetime. When the Van de Graaffmachine came along, it providedreliable 2 million volts, or more,energy, and X rays became extraor-dinarily penetrative that could pene-trate steel plate many inches thick.

Recently, there was a very largemachine produced in Europe,designed for over 20 million volts,to meet the needs of researchersinterested in using X rays to altermaterials; radiation can be a veryeffective cross-linking agent toimprove the properties of materi-als. An example of work MIT hasdone in cooperation with Massa-chusetts General Hospital involved

The Van de Graaff was constructed on railroad track bases so that it could bewheeled out of and within the hangar.

CO

UR

TE

SY

MIT

MU

SE

UM


13

IEE

E IN

DU

STR

Y A

PP

LICA

TION

S M

AG

AZIN

E •

JAN

|F

EB

20

05

• W

WW

.IEE

E.O

RG

/IAS

the use of radiation to improve thematerial used in hip replacements.Before that development the mate-rial used for hip replacementwould last 10–15 years, and thenthe patient would expect to under-go an operation for a new replace-ment. With the present use ofradiated material technology, hipreplacements appear to be perma-nent, not wearing out in one’s life-time. There are many otherapplications where radiation is avery effective application. Plasticsof many kinds are cross linked tomake them stronger, more durable,and operate at higher temperature,and organic materials can be madeelectrical conductors. Semiconduc-tor devices can be irradiated toimprove their operation [2].

In machines in which the high-voltage insulation is air, breakdown can be frequent, and thesparks are typically 10–20 ft long.In an exhibit at the Museum of Sci-ence, in Boston, the audience isprotected by a series of wires thatrun vertically in space, about 6-inapart, and the screen acts effectivelyas a Faraday shield that isolates andprotects the audience from the elec-tric fields. The theater was createdfor it, and the grounding system forit is quite significant.

Going to compressed-gas environ-ment quickly became the way to getmuch higher voltage; several high-voltage machines were built at MITand other institutions. At the Uni-versity of Wisconsin, Ray Herb pro-duced a compressed-gas insulatedmachine of the Van de Graaff typethat was quite successful. Both Vande Graaff and Ray Herb ended uptransferring their technologies toindustry for commercial use of elec-trostatic machines. The machines atMIT were used for physics research,their first target because it was VanGraaff’s love and the motivation fordeveloping the technology was origi-nally for nuclear physics, and laterrecognized they could be used forradiation to improve the characteris-tics of materials.

Miniature Van de Graff GeneratorsSome companies in the UnitedStates make miniature machines for

education and entertainment pur-poses. Mr. James Davis, of theWabash Instrument Company ofWabash, Indiana, told the writerthat over 15,000 miniature Van deGraaff generators have been sup-plied by his company to schools inthe United States. They are desk-size and used for teaching physicsand chemistry. The small machineswill produce up to 150,000 V. Themaximum continuous current isonly about 10 µA, so the units arequite safe; Sparks of 8-in long maygo to 12–15 in [3].

The Westinghouse Atom Smasher [4]The Van de Graaff generator built byWestinghouse in 1937 at its researchfacility in Forest Hills, Pennsylvania,was constructed for the first nuclearphysics program out in industry. It wasintended for an unprecedented highvoltage. The machine was designed bya group of Westinghouse engineers ledby Dr. William H. Wells. Wellsreceived the Ph.D. degree at JohnsHopkins University in 1934; he hadbeen on the faculty of the University ofMinnesota before joining Westing-house Research on June 1936.

The design drew on the experi-ence with pressurized machines byBarton, Mueller, and Van Atta atPrinceton, and Herb, Parkinson,and Kerst at the University ofWisconsin. The pressure tank is30-ft in diameter and 47-ft high,and the top of the machine is 65-ftabove ground level. The air pres-sure of 125 lb/in2 was used toincrease the breakdown voltage forthe internal sphere and column.Since the voltage was distributedalong the column, lesser standoffdistances were required at lowerportions, which accounts for thepear-shaped design of the tank.

The 30 ft-long accelerationvacuum tube consisted of a stackof porcelain rings with 12-ininside diameter, interleaved withmetal spacers that formed theaccelerating electrodes. At the topof the accelerating tube, support-ed by four columns of porcelaininsulators, is a spherical shell con-taining the ion source, a belt-dri-ven generator for power, and setsof needles to charge the two main

belts that traveled along theassembly. The inside of the tubewas evacuated. At its lower endwere magnetic analyzers to sepa-rate the accelerated particles intothose of the proper charge andmass. The bombardment of tar-gets took place on the groundfloor, well below the ground levelif extra shielding was desired.

By July of 1937, the tank hadbeen erected, major internal struc-tures lowered into place throughthe open top, and the tank closed.By this time, the company felt itwas appropriate to let the worldknow what it was doing. Wells wascautious about the anticipated per-formance, as both MIT and theCarnegie Institution were havinggreat difficulties obtaining highvoltages with their newly con-structed Van de Graaff generators.The machine was accordinglydescribed to the press as a five mil-lion volt machine, and so it hasbecome known. Although it didrun above 4 million volts, it neveractually achieved five. The strikingappearance of the machine, thesuperlatives of voltages in the mul-timillions, atoms of incomprehensi-ble smallness traveling atunimaginable speeds, and theglamour of penetrating into the sci-entific unknown, all made goodcopy, and newspapers and maga-zines published good accounts andpictures of the machine.

In 1984, the machine was nomi-nated to be designated an IEEEMilestone in Electrical Engineering;it was so designated on 29 May 1985.

An example of the mini Van deGraaff generator.

CO

UR

TE

SY

WA

BA

SH

INS

TR

UM

EN

T C

OR

PO

RA

TIO

N


IEE

E I

ND

US

TRY

AP

PLI

CA

TIO

NS

MA

GA

ZIN

E •

JA

N|

FE

B 2

00

5 •

WW

W.I

EE

E.O

RG

/IA

S

14

The Westinghouse Atom Smasherstill exists, though its operation wasdiscontinued around 1950 when itwas replaced by a commercially builtmachine. All buildings of the origi-nal Westinghouse Research facilitybuilt in 1916 have been demolished;the Westinghouse atom smasherstands alone. Westinghouse builtand moved to much larger and mod-ern facilities in 1954. The presentowner of the site is trying to disposeof the property, the ultimate fate isstill to be determined.

Early History of Electrostatic Machines [5]The Van de Graaff generator is thelast in a long series of machinesthat date back to the earliest timesin the history of electricity. Ourword electricity comes from theGreek word for amber: “ilektron.”In 600 B.C., Thales noted thatwhen amber was rubbed, it wouldattract light objects, and in 1600A.D., William Gilbert, in hisepoch-making treatise on magnet-ism, was the first to use the term“electrical” in its modern sense.

The generation of electricaleffects by friction was the basis for anumber of electrical machines.Otto von Guericke (1602–1686),known for his famous experimentin which teams of horses could notseparate the halves of an evacuatedsphere, used a ball of sulfur turnedby a crank and rubbed by the hand.Isaac Newton substituted a glasssphere for the sulfur ball. Furtherimproved machines using glassdisks and Leyden jars to store moresubstantial amounts of charge wereimportant tools for describing thevarious properties of electricity.

Benjamin Franklin, one of thegreat pioneers of electrical science,used such a machine and providedan improvement that is one of thekeys to the Van de Graaff genera-tor. He showed that pointed nee-dles, in close proximity to acharged body, would conduct copi-ous amounts of electricity. He usedthis in the research that culminat-ed in his famous 1751 demonstra-tion that lightning was anelectrical phenomenon, and thatelectricity derived from his kitestring had all the properties of

electricity generated in the labora-tory by the friction machine.

Somewhat later, a quite newclass of electrostatic generatorscame on the scene. These “influ-ence machines” transferred chargefrom a source having low potentialto one of higher potential. Theprincipal is easily understood withthe help of some elementary equa-tions. A capacitor with capacitanceC and charge Q will have a voltageV = Q/C. If the capacitor is madeof two parallel plates, and thoseplates are separated, the capaci-tance C will diminish, and sincethe charge Q is constant, the volt-age V will increase. Now, the twoplates of opposite charge attractone another, and in order to pullthe plates apart, work must bedone. The work done in pullingthe plates apart appears as higherelectrical energy. The charge car-ried by the separated plate isdumped conductively into anoth-er capacitor. The plate is then car-ried back to be close to the firstplate, momentarily grounded, andthe process repeated. In many ofthese machines the charge is fedback to the first plate, so the twoopposite capacitor potentialsclimb up one after the other toquickly achieve values limitedonly by leakage or breakdown

Machines based on this princi-pal first appeared in 1787, andover the years took a variety offorms, all involving some means ofseparating charged conductors, anddischarging them conductively asappropriate. The most highlydeveloped was the Wimshurstmachine of 1878; it consisted oftwo closely spaced glass disks on acommon axis, rotating in oppositedirections, and carrying sets ofmetal foil carriers playing the roleof capacitor plates. Conductorsintermittently touching the foilcarriers took the charges to thepositive and negative terminals.That machine, having a substantialoutput at high voltage and beingquite reliable, was widely used as avoltage source for X-ray machines.An elderly emeritus professor inthe physics department at the Uni-versity of Illinois was still using hisWimshurst X-ray machine in

1938—he saw no reason to updateit as it worked fine. Interestingly,another feature used later in someVan de Graaff generators was intro-duced in 1900—the enclosure ofthe entire machine in a chamberthat could be pressurized with airor carbon dioxide, enabling it toproduce voltages as much as threetimes those obtained in open air.

The Van de Graaff generator isan influence machine. The capaci-tor plates are no longer easily visu-alized, as they are continuouslydistributed as the surface of thebelt itself. Charges are carriedmechanically by the belt anddeposited on the high voltage ter-minal. The electrical energy thusderived comes from the mechanicalenergy put in to driving the belt.

The writer extends sincere appre-ciation to Drs. Chatham M. Cookeand John W. Coltman for their con-tributions and assistance in generat-ing this history article.

References[1] American Heritage Dictionary of the English Lan-

guage. American Heritage Publishing, 1970.

[2] Various information from MIT Van de Graaffhistory Web sites.

[3] Wabash Instrument Corp., Wabash, IN.

[4] J. Coltman, “The Westinghouse atomsmasher—A IEEE historical milestone,” IEEELog Number 8610623, Nov. 4, 1985.

[5] Encyclopedia Britannica,11th ed. 1911. IAS

The Westinghouse Atom Smasher.

©W

ES

TIN

GH

OU

SE


Documents

A history of the Van de Graaff generator