Tooth Tables: Solution of a Dental Problem by Vector Algebra

Author(s): Gary Hosler Meisters

Source: Mathematics Magazine, Vol. 55, No. 5 (Nov., 1982), pp. 274-280
Published by: Mathematical Association of America
Stable URL: http://www.jstor.org/stable/2690094 .
Accessed: 18/06/2014 03:38

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to
Mathematics Magazine.

http://www.jstor.org

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
ToothTables: Solutionofa Dental
ProblembyVectorAlgebra
GARYHOSLERMEISTERS
ofNebraska-Lincoln
University
Lincoln,NE 68588

The following problemin dentistryconcerning thewayin whicha toothis preparedfora gold

inlaywas posed to me by twoinstructors at theUniversity of NebraskaDentalCollege,Linda
Dubois and StanleyKull. Theyexpectedthata mathematical analysisof theproblemwould
requireheavyuse of a computer. However,the problemtheyposed is nicelyamenableto a
formulation (and solution)usingsomeelementary factsofthree-dimensional vectoralgebrawhich
are usuallyincludedin thestandardfirstcoursein calculus.The solutionwe obtainedand its
implications fordentistryhavebeen describedin thepaper[2]. But,since[2] was written for
themathematics
dentists, wasonlybriefly summarized andlittleattempt was madeto explainthe
detailsof themathematical formulation and solutionof theproblem.Consequently, it seemed
appropriateto writethispaperin orderto givea fullaccountofthemathematics involved.
In orderto describetheproblem wewillneeda smallamountofdentalterminology whichcan
be easilylearnedfromFIGURE 1 and the accompanying "Dental Vocabulary"gleanedfrom
Dorland'sMedicalDictionary [1]and Gray'sAnatomy [3].A toothis preparedfora goldinlayby
carvingout(fromtheocclusalsurfaceand froma proximalsurface ofcontactwitha neighboring
tooth)a cavityshapedmoreor less as illustrated in FIGURE 1. This cavityis called an inlay
preparation.
Theproximal boxis thatpartoftheinlaypreparation whichis boundedin therearbytheaxial
wall,on thesidesbytheleftandright walls,on thebottombytheseat,on thetopbya portionof
theocclusal(whichhas beencutaway),and on thefront by a portionof theproximalsurfaceof
contact(whichhas also beencutaway).

Occlusal
surface

Buccal
surface -

~~~~~~~~~~surface

Leftwall of wall l Rightwall of

theproximalbox Seat theproximalbox
(thebuccal wall) (thelingualwall)
Proximalsurface

FIGUR 1. A tooth(molar)showinga gold-inlay carvedout of it.

preparation

274 MATHEMATICSMAGAZINE

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
 DentalVocabulary
A Brief

is thecheeksideoftooth.
Buccalsurface

Lingual
surface sideofthetooth.
is thetongue

is thechewing
Occlusalsurface surface
ofthetooth.

Preparation is theshapedcavitycutout of a toothto receivea goldinlay.

(or inlaypreparation)

Proximalis nextto theneighboring

tooth.

nearestto theneighboring
Proximalbox is thatportionof theinlaypreparation tooth.

Proximalsurfaceis thesurfaceof thetoothwhichis in contactwiththeneighboring

tooth.

Seat (or gingivalseat) is theseat of theproximalbox nearestthegums.

Floor of inlaypreparationis parallelto the seat but is closer to the occlusal surface.It is
more-or-less
perpendicular to theaxial wall and extendsfromtheaxial wallintothecenter
ofthetooth.Theproblem in thispaperis notconcerned
discussed withthepartofthe
which
inlaypreparation liesbetweenthefloorandtheocclusalsurface.

It is required to cutthecavityforthegoldinlayin sucha mannerthata waxmoldofthecavity

can be drawnouttowardtheocclusalandthegoldinlayitselfcanbe droppedintothecavityfrom
theocclusalsideofthetooth.Thisis calledan occlusaldraw.In orderto guarantee a goodocclusal
drawitis bothnecessary and sufficient
thatall ofthenearly-vertical wallsoftheinlaypreparation
shouldbe tiltedslightly backso thattheyall faceslightlytowardtheocclusalsurface ofthetooth.
The linesofintersection oftheaxialwallwiththeleftand rightwallsoftheproximal box will
be calledtheaxialcornerlines.Theircontinuations aremarkedin FIGURE1 by arrowsB and L
(for"buccal" and "lingual").Sincetheseaxial cornerlineslie in a commonplane(namely,the
plane of the axial wall) theyare eitherparallelor else theyintersect. If theyintersecton the
occlusalsideoftheseat,thenwe saythattheselines"converge in thedirectionoftheocclusal"or
simply"converge." On theotherhand,iftheyintersect on therootsideof theseat,thenwe say
thattheselines"divergein thedirection oftheocclusal"or simply"diverge."
It hasbeencustomary fordentiststo cutinlaypreparations in sucha waythattheaxialcorner
linesdiverge in theocclusaldirection bythetwoarrowsB andL shownin FIGURE1.
as illustrated
Evidently thisdivergence of B and L has been generally regardedby dentists(see [2] and the
references citedthere)as bothnecessaryand sufficient in orderto avoid an undercut by the
proximalbox wallsin an inlaypreparation. Indeed,divergence used as a visual
is frequently
criterion fora good occlusaldraw.It is themainpurposeof thispaperto proveby meansof a
mathematically rigorous argument thatthisdivergence is neither necessarynorsufficientforan
occlusaldraw.In factwe showthat
ofB andL toward
(i) Convergence theocclusalcanbe consistent
witha goodocclusaldraw;
is noguarantee
(ii) Divergence ofa goodocclusaldraw.
One consequenceof thisresultwhichmayhave important significancefordentistry
is that
whenan inlaypreparation is cutwithconvergence withdraw)it allowslessgoodtooth
(consistant
materialtobe cutawaythanwhenone requires divergence.
It remainsof coursefordentalinstructors to perfecta procedurewhich,on theone hand,
allowsconvergence and guarantees a gooddrawand,on theotherhand,is simpleenoughto be
to learnand to be able to putintopractice.In thispaperwe havemerely
practicalfordentists
shownthatit is geometricallypossible.

VOL. 55, NO. 5, NOVEMBER 1982 275

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
 and solutionof theproblem
formulation
The mathematical
We are concernedwiththe geometry of the proximalbox, and it sufficesto considerthe
right-hand sidesincetheleft-hand sidecouldbe treatedin a completely analogousmanner.We
introducea three-dimensional rectangular coordinatesystemorientedso thatthe seat of the
proximal box liesin thexy-plane, thez-axiscoincideswiththecentralaxisof thetooth(directed
fromtherootoutthrough theocclusalsurface ofdraw),they-axisis parallelto the
in thedirection
planeoftheaxialwallandpassesoutthrough and thex-axispiercesat
theright(lingual)surface,
rightanglesthelineofintersection oftheaxialwallwiththeseatand thenexitsthetooththrough
theproximalsurface.This rectangular coordinate in FIGURE 2. Insteadof
systemis illustrated
comparing therightL and leftB axial cornerlines(forconvergence or divergence towardthe
occlusal),we willcomparetheright(or lingual)axialcornerlineL withthelineofintersection of
theaxialwallwiththexz-plane.Thislatterlinewe call the"medial"line.
As usual,we use theletters i,j,k to denotetheunitvectorspointing in thex, y,z directions,
respectively.Recallthattheorientation spacecan be conveniently
ofa planein three-dimensional
describedmathematically by meansof a unitnormalvector(a unitvectorperpendicular to the
surfaceof theplane).In FIGURE 2, thevectorlabeledA is theunitnormalvectorto theplaneof
theaxialwall,and thevectorlabeledP is theunitnormalto theplaneoftheright(lingual)wallof
theproximal box.Bothofthesevectorsarechosenso thattheirdirections pointintotheproximal
box.Alsoin FIGURE 2, thevectorlabeledM is parallelto themediallineand thevectorlabeledL
is parallelto theright(lingual)axialcornerline.
For theformulation andsolutionofourproblem we shallneedtouse thefollowing well-known
properties of the"cross"and "dot" products of thetwovectorsU and V, whichcan be foundin
anygoodcalculustext.
(1) Definitions.If U=uli+u2j+u3k and V=v1i+v2j+v3k, thenthecrossproductof U
and V is definedas
i j k
UXv= Ul U2 U3
VI V2 V3

or
U XV = (u2V3 - u3v2)i- (uv3-u3v)j + (uv2 -u2vI)k.

z-axis
ofocclusal)
(direction

Floorofocclusal
Mt t inlaypreparation

Medial I i Rightaxial
Medial Icorner 4 line

?
Axialwall /)/ (out through
lingual)

A x-axls~
/ ~ ~~y-xi
wallofthe
\\i_~~~Right

proximal)
(outthrough
FIGURE2. Anappropriate coordinatesystem.
rectangular

276 MATHEMATICSMAGAZINE

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
Thedotproduct ofU andV is definedas U V = uIv1+ u2v2+ U3V3. ThelengthIUI is definedto
be [(U1)2+ (u2)2 + (u3)21/2.
(2) U X V is perpendicularto bothU andV.
(3) UXV=-VXU.
(4) The "right-hand rule."If 0 < r is the(smaller)anglebetweenU and V, thenU X V
pointsin thedirection of thethumbwhentheright-hand fingersarefolded(throughtheangle+)
fromthedirection ofU intothedirection ofV.
(5) JUXVj=jUH Vjsin+.
(6) U.V=iUjIVIcos4.
(7) UX(VXW)=(U-W)V-(U-V)W.
ul U2 U3
(8) U (V X W) = (U X V) W = vi V2 V3 . thetriple
If thisproductis positive(negative),
WI W2 W3
U,V,W is calleddextral(sinistral).
Now notethatsinceM is parallelto thexz-planeand also to theaxialplane(see FIGURE2), M
to both of the vectorsj and A. Therefore
mustbe perpendicular (since the lengthof M is
we maywrite
immaterial),
M=A>Xj, (9)
and similarly,
L=PXA. (10)
SinceM and L arebothparallelto theaxialwall,it followsthateitherM X L = 0 (whenthey
areparallelto eachother)orM X L is a nonzeroscalarmultiple of
ofthevectorA. Consideration
theright-hand ofA yieldsthefollowing
rule(4) and thealgebraicsignof thisscalarcoefficient
criteria:
I. Theright lineconverges
(lingual)corner themediallinein thedirection
toward oftheocclusal
when
precisely
MXL=XA, withX>0.
II. Theright(lingual)cornerlinediverges oftheocclusal
fromthemediallinein thedirection
when
precisely
MXL= -XA, with
X>0.
fora goodocclusaldrawis met(as far as theproximalbox is concerned)
III. The condition
whenthenormal
precisely vectors
A andP (as wellas thenormalvectortotheleftwall)each
havea positive
z-component.
The questionthenis this:Is occlusalconvergenceof L towardM consistentwitha good
occlusal draw?That is, can conditionsI and III be simultaneouslymet,ratherthan (as
onlyII and III?
believedbydentists)
traditionally
In orderto answerthisquestion,we willneedthex, y, and z components
of theunitnormal
vectorsA and P. Thesecan be obtainedfromFIGURs 3, 4, and 5 whichare two-dimensional
extracted
figures fromFIGuRE 2.
FromFIGuRE3 we see that
A = (cos a)i + (sina)k, (11)

x
cos a
FIGuRE of thevectorA; here6= sina.
3. Components

VOL. 55, NO. 5, NOVEMBER 1982 277

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
 y

Base of therightwall
of theproximalbox
P

xy-plane |
cos x
E of thevectorP; hereE
FIGURE 4. The z-component sin4. FIGURE5. Projectionof P ontothexy-plane.

and fromFIGUREs4 and5 weseethat

P = (cos 0 sina )i-(cos 0 cosa )j + (sin
')k. (12)
Extractingthez-components from(11) and (12), criterion III saysthatthecondition forgood
occlusaldrawis metwhen8 = sina > 0 andE = sin0 > 0. Theaxialplanewillbe undercut if8 < 0,
and therightwallof theproximalbox willbe undercut ifE < 0.
We can applysomeof theproperties of crossproductto rewrite M X L in criteria
I and II:
M X L = MX (P X A) = (M -A)P - (M. P)A =-(M P)A.
I and II can be rephrased
Thuscriteria as follows:
It,II. M P shouldbe negative
forocclusalconvergence
andpositive
forocclusaldivergence.
(8), (9), (11), and (12) we obtain
Butbyformulas
cosa 0 sina
M P=(AXj).P= 0 1 0
cosp sina -cosp cos a sin
= cosa sin -sin a cos0 sin0.
thecriterion
Therefore, oftherightaxialcornerlineL towardthemedial
forocclusalconvergence
lineM becomes
tan < tana sin , (13)
fora goodocclusaldrawmaybe expressed
whilethecriterion as
tan >O and tana>O, (14)
becausecos0 and cosa arepositivein therealistic
rangeI0 1< r/2 and I <<r/2.
Conditions (13) and(14) mayeasilyholdsimultaneously ofvaluesofa,
fora continuum , and
0. Thatis, convergence consistent
is definitely witha goodocclusaldraw.
The reverse inequalities
tan+> tanasin0, (15)
and
tan0 < O and/ortana < O, (16)
are also easilysatisfied
simultaneouslyfora continuum ofvaluesof0, a and 0 with0 < 0 < 7T/2.
But (15) is theconditionforocclusaldivergence of therightaxial cornerlineL awayfromthe
mediallineM, and (16) is thecondition forfailureof occlusaldraw,becauseof an undercut of
eithertherightproximal wall(tank< 0) or theaxialwall(tana < 0). Of course(15) impliesthat
tan0 can be negative onlyiftana is also negative,becausesin0 is alwayspositive(O< 0 < 7T/2).
Thusocclusaldivergence is nota guaranteeofa goodocclusaldraw,and thusshouldnotbe usedas
a visualtestforsuch.Note thatforsimultaneous convergence and drawit mustbe truethat
tan0 < tana because0 < sin0 < 1.

278 MATHEMATICSMAGAZINE

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
 The criterion(IF) thatM -P shouldbe negativeforconvergence has important geometric
The dotproductM- P negative
implications. meansthatthecomponent of P in thedirectionof
M is negative, ofP alongthelinethrough
i.e.,theprojection M has a directionoppositeto thatof
M. In addition,M P =(A Xj) -P is negative if and onlyif thethreevectors{A,j,P} forma
left-handed
(sinistral)
coordinatesystem.
A tableofvaluesfortheanglesthatallowsimultaneous
convergence
anddraw
In additionto theanglesa, 4, and 0 introducedin FIGURES3, 4, and 5, we can introducethe
angleg betweenM and L and call it the"angleofconvergence" (takenpositiveforconvergence
fordivergence).
and negative We definedM in (9) as M = A X j; usingequation(I1), we findthat
M = A X j = (-sin a)i + (cos a)k,
so thatM is a unitvector.From(6) it followsthatM *L = LI cos,u.Also,from(11) and (12) we
have
M L = (A X j) *(P X A) = cos
4 cos0,
so that
cos,u= (cos4cos 0)/I LI. (17)
(3), (5), and (8) giveM X L = IL IA sin,u= -(M *P)A, so that
Properties
sin,u= (sina cosOsin0-cos a sin )/ILI. (18)
(17) and (18) gives
Combining
4 sec0.
tan z= sina tan0-cos a tan (19)
Thustheangleofconvergence p.is completely
determined by theanglesa, 4, and 0.
By usingstandardtablesof the trigonometric functions
(such as thosefoundin [4]), or a
calculatorwithtrigfunction a tableofacceptablevaluesofthe
onecan easilyconstruct
capability,
anglesa, 4, and 0 and computethecorrespondingvaluesof,ubymeansof theformula (19). We
givean exampleofsucha tablein TABLE1.

a 4, 80~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
60 30 300 0.01l
,, ,, 350 0.550
,, ,, 400 1.130
,, ,, 450 1.770
1,',, 5QO 2.490 a = angleof tiltof axialwallback fromvertical.
550 3.340 = angleof tiltofproximalwallback fromvertical.
600 4.4390
650 5.760 8 = angleof (tangentplane) of rightwallof
700 7.680 proximalbox.
70 30 260 009?u = angleofconvergence
betweenvectorsM and L.
,, ,, 30? 0.590
II 1/ ~~350 1.250
400 1.970 O<tan4, <tanasin |
450 2.770 Draw t t Convergence

500 3.680
,, ,, 550 4.770
,, ,, 600 6.110
650 7.870
700 10.36?

TABLE 1

VOL.55, NO.5, NOVEMBER1982 279

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions
 I wouldliketo thankPaulCampbell meto expandandwriteup thisproblem,
forencouraging whichwasthe
topicofa talkI presented
totheNebraska-South
DakotaSection
oftheMAAat itsmeetingin Vermillion,
South
DakotainAprilof 1981.

References
[1] Dorland'sIllustratedMedicalDictionary,Twenty-fifth
ed., W. B. Saunders,Philadelphia,1974.
[2] Linda Dubois, Gary H. Meisters,and StanleyKull, The relationship of wall orientationto proximalbox
(submittedto theJournalof Dental Research).
designin an inlaypreparation,
[3] HenryGray,Anatomyof theHumanBody,Twenty-ninth AmericanEdition,C. M. Goss, ed.,Lea & Febiger,
Philadelphia,1975.
[4] S. M. Selby,ed., StandardMathematicalTables, Twenty-first ed., The ChemicalRubber Co., Cleveland,
1973.

Periodic Sums of Functionswith

IncommensurablePeriods

REGINA CAVENY
JAMES CAVENY
ofSouthern
University Mississippi
Hattiesburg,
MS 39406

Iff and g are realfunctions

withfundamental periodsa and fi,respectively, whenwillthesum
f + g beperiodic? Variationsofthisquestionarisenaturally as soonas theconceptofperiodicity is
introduced, perhapsin a trigonometry course,and surfacerepeatedly in moreadvancedmathe-
maticaltreatments whichinvolveperiodicfunctions and theirapplications to variousphysical
phenomena. If we admitonlycontinuous functions, thenthesumf + g is periodicif and onlyif
theratioa/13is rational.But evenforcontinuous functions theproofsare elusive.Thispaper
deals mainlywiththequestionof whether or not theirrationality of a/,8 insuresthatthesum
f + g is notperiodic.We givesomeexamplesto showthatthesumcan be periodicwhena/13is
irrationaland can evenhave a fundamental period.Thentwoconditions are givento establish
thatthisdoes nothappenforthefunctions whicharenormally encountered in applications. The
arguments are, forthemostpart,self-contained and accessibleto any seriousundergraduate
studentof mathematics. We makeno claimof originality of results,whichmustsurelybe fairly
well known,but whichseemto have escapedthepopularliterature. Our intentis simplyto
assemblea reasonably completeinvestigation whichwillbe accessibleto thegeneralreaderofthis
Magazine.
For preciseness we agreeon thefollowing terminology andnotation. Throughout thepaper,J
denotesthesetof integers and R thesetof realnumbers. A function f definedon R is periodic
withperioda iff(x + a) =f(x) forall x in R. If thereis a positiveperioda suchthata ? fiforall
positiveperiods,Boff, thena is calledthefundamental periodoff. Two nonzerorealnumbers a
andf8arecommensurable iftheirquotientis rational(i.e.,thereexistnonzerointegers i andj such
thatia +j/ = 0) and areincommensurable iftheirquotient is irrational
(i.e.,forall integersi and
j, ia +j,B= 0 impliesthati =j = 0). IfA and B are subsetsofR, thenwe define
A +B {x= &tR: x a+bwitha &A andb&B}
AC={xER: x A}
and,forr E R,
rA= {x E R: x = rawitha EA}.

280 MATHEMATICSMAGAZINE

This content downloaded from 176.31.127.168 on Wed, 18 Jun 2014 03:38:11 AM

All use subject to JSTOR Terms and Conditions