NEWS LETTER L2Bitish Technial Committee, Corpw Vitrcatum Medii Aevi

Department of Physics. University of York l0 January 1975 Ref. TC/75ll

b7$2-GIS#'aJ'j1rh4 R* L,*-r,.rJ ^^t !.,ot l, r $;r\\ u*rrg rc..rt-\., tt*CONTENTS

1 GENERAL

2 DOES CLEANING AFFECT WEATHERING ?

3 HEATING OF MEDIEVAL STAINED GLASS !

4 PROTECTIVE GLAZINGS

5 NEW ABSTRACTS

t

l GENERAL

1.1 PROFESSOR IIANS HAIINLOSER

The death of Professor Ha.ns Halrnloser on7th November L97\, after a protracted- and verypainful illness, has robbed the CVMA (and. alsoCfIIA and other organisations ) of an inspired.and energetic leader. His passing has caused.great sorrov in aIL quarters a.nd it vill bed,ifficult to replace him. Our d.eepestsympathy goes to his brave wife anil tbe restof their fanily.

I unilerstand that Professor LouisGrod-ecki iri11 tenporarily assume the dutiesof Acting President of the CVMA until a newPresicient can be elected.

7.2 E)OIIBITION OI'YORKIS CONTRIBUTION TO

STAINEI GLASS

Reaclers of these l::::::::::::::::Ievs Letters in GreatBritain nay wish to make a note in theirdiaries that an irportant exhibition vi,l-l- bemounteci in York during the week beginningJth April 1975, on the occasion of an Inter-national Congress to be held in the Universityby the Chenical Society. The erctribition isexpected to be staged. in the foyer of theDepartment of Physics and it vill cover fourdifferent aspects of Yorkrs contribution tostainecl glass:-

( a) Photographs of importent vindovs in theI{inster, and in some of the churches in york,and actual panels from the Minster both beforeand after restoration.(U) Examples of the conservation activitiesof the York Gl-aziers Tn:stn probably to beshorrn as fil-n on W monitors in the exlribitionafe ad

(c) Exptanatory exa.mples of the part played.by the University in iconographical researchvhich assists the proper restoration of awind-ow. Dr Peter Nevtonrs stu{y of theAngels window at A11 Saintsr Church, ltrorthStreet vil-l be one of the exa,rnples,

(a) ExrmFl-es of the scientific vork carrieclout by the University on the chenicaJ_ compos-ition of the Minster gJ-ass in rel-ation to itsremarkably varied durability patterns. Otherexamples of scientific aid"s to conservationvi1l be displayed, such as the recovery oflost inages, the use of radiation nonitoringto establish the authenticity of pieces ofear\r med-ieval glass, and a studtrr of thehr:midity existing around the old- glass vhenprotective glazing is used..

Ful-].er d"etails of this exhibition, '^'ithtimes of opening, vil-l be given in laterNevs Letters.

r.3 PAUL JETFERIESI REPORT

In Nevs Letter lto.6 (itern 1'1"'1) it vas

stated that Mr Paul Jefferiesn of Mr Dennis

Ki;;;- vori<shop in Norwich, hact-been arsarded

tfr""ft..r.U-ing Schotarship for Stained Gl-ass

oifered by bhe Rad'cliffe Trust $cheme for the

irafts. He mad'e the journey betveen September

13th and October 5ttr ana visited Rouent

ui"",o, Paris, Sens, . Chartres, Troyes '

sttr"uontg, Ufo, Munich, Vienr:a, -Regensburgt

Altenburg, Augsburg, Straubing, Nurembergt

Linnich, Cologne, Aachen and Goud'a'

At the laboratory at Champs-sur-Marne he

sav glass fragments cleanecl by M' Betten-;;;;;;;';;;;; (see $'L' No'7r item 2'r+);

lr." Er""t was matt ancl translucent but iti*ptJ""a vhen coatecl' He a1so. sav the

r-"ili"" referrecl to in iten 159 on P'? ofN.L. No.11; the resins used' for coatingst"i"t""a to in item t'JO; antl the use ofvi"."yr vC:53 referred to in the same item

;;;;-"f;" seating freshly-cleaned surfaces and

fi-xing loose Paint'

In Munich, at the Frauenkirche' much oftrr" oia gf;"" *ttj'"n vas clea^ned in the period

ig5o-eo i" ,ro* said to be badly corrocled'Tne Sctrarfza^nclfenster vas in better conditionthar most.

At St Maria am Gestad'e in Vienna he

noti-ced that sone light penetrated betveen

tir" tiAettea medievaf panels and the.sides ofii,u *nrfio"s in the isothe:rnal glazing

"V"t"t there, but it vas visible only from

the sanctuary area'

In Nuremberg, at the vorkshoP ofDr Frenzel, he sav the restoration of the

ArguUr."g Prophet windovs, any loose paintf'"i"e-rl""a iritir Aralclite AYlol' The backs

of tf,e vinclovs vere vashed vith distilledvater and any loose cortosion vas scrapecl

avay vitn a scalpel vhile it vas still vet'n""iail", thinne-d vith acetonet vas used- forii*i.tg any loose paint ' r'ractures vere ed"ge-

l""a"i wiin nrato:ite ancl- any gaps were filledvitfr ,trafaite and afterwards toned dovn vitha suitable eold Paint'

Dr Frenzel- made a statement that' in the

l-9th Centuryt sone meclieval glass had- been

r"puintua an6' refireA' lnCu - see afso iten 3

belov. ) Dr Frenzel is aJ-so experimenting vithefectrieaffy-heateci- copper vires in the cavityoi-an i"otnlrmal glazing systent to,keep the

interspace I degCl *ar*ut than the interior ofiir"-u"iral"e. -(ncu - it vilI-be importaat to

lealn more about this interesting experimentl.;;;;* the vires are separatedl what currenti-s use,1; ancl hors the temperature rs con-

i"oir.a. ) I,lhen the exterior appearance of ari

isotnernaf system is important (as at,tri""u"tgt tle outer glazing is leaded- to copy

It" **itt'reaci-lines of the medievaJ- glass'

At Cologne he visited both the Dom and

St Kunibertts Church ancl helped to clear up a

critici-sn of the rJacobi proeessr' At

St Kunibertrs there vas much d'iscolouration

"i tit" bontled joints of the St Catherine

vindov due to the use of an earlier technique'

fn the Dom a revisecl process is in use and no

discorouration has occurred.. (ncn - A partyof art-tristorians and scientists who vent toCologne and san only St Kuni-bertrs have been

."p-t;y-""iti.caJ- of trrl Jacobi process and itt".uG that their viev may now need to be

"""i""a, but a photograph of the St Kunibertfs

;i"e;" i" ,rot availabre for reproduction inthe News tetter.)

1.I{ STORAGE OF VAI.UABLE GLASS DURING THE WAR

Miss Judith Scott, late1y of the Councilfor Places of Vorship, has kindly clravn

attention to the existence of var-timerecords regarding the storage of religioust""*ll"u" of *ff kinds. Many of these, inc-1ud:ing staineil glass from Exeter, SalisburytTickenham, lJeston in Gord'ano aricl l{altont 1^rere

storecl in the crypt of Viveliscombe church inSomerset, anil some church treasures vereapparentiy damaged- by the floocl.vhich occurrediiu". i" i9\7. -ttt""L records will neecl timefor study but Mr John Snrith has dravn ryattention to a parq:hlet issued in f9\O by theCentra^L Council for the Care of Churches

""liir"a ttHov to protect a church in var timett'

on page l-3 it states that panels of glasswhi.ch have been removed. shoufd ttbe tigfrtlypackecl in strong boxes or crates vith drysavdust betveen each panel' The crates shouldbe removecL to a safe place and stored so that;;"-;;;"i" are in an uprieht position"'

Even though ttd-ry sa"rdusttt rsas reconmerldedt

it wou.l-d. not remain ttd-rytt for very long' For-*ampfe, in a cell-ar vith a teuperature ofidtC;a a ret-ative hr:nidity of 90% the saw-

dust vouad probably have a moisture contentof 28i1, (aeplnaing somevhat on the type of r'rood'

fron vhich the savdust vas mad'e)' Fortunately'moist sar,rdust tends to be sliglrtIy acid(rather than al-kaline vhi-ch vould tend todamase the glass), Scots pj-ne havi-ng pH valuesor l+l: to l+.6, spruce l+.8 to 5'0 and beechI+,5 to 5.9a oak is distinctlv acia (3'3 to:.g) lut ii voul-d probabty not have been used

for sarrclust. Hoveverr the question should be

investigated in an experiment at the YorkGlaziers Trust. Some si-nulated med-ieva1 glass(Sritish Glass No.2) vill be packed in savdust

and stored in a damp cellar to see vhether any

da.mage occurs .

1.5 DEPOSITS INSIDE THE PLATING OF THE HEAD

AT CANTERBURY

.fn Section 2 of N'L' No'10 an account vas

given of the condition of the moclerrr glass

fnside a platect head at Canterbury Cathedra'I'In the fourth paragraph of p'2 it vas stated.

that Dr N.J. Brid-ge voul-cl be studying thevhite and black deposits found on.it' He has

nov reported' that the vhite material vas

rypsun and the black material vas probably a

il.-*tn"" of graphite and- silica geI' The

graphite presu-lrabfy came from soot- in theitnispnere. The glfpsun anci the silica gelrightalso have been airborne (eg from stone-*-it, .t".) but nay have come from glass, invhich case they vould- have come from thea^ncient glass because the moclern glass vas

shinY and unaffected''

2 DOES CLEANIT\IG. AFFECT WEATHERING ?

There are suggestions in the }iteraturethat the erust, or the patina, on weathereclned.ieval glass may have a protective actionagainst further weathering (see, for exampletAbstract ttos. l+5, l+p ancl 5o in the BritishAcaclemy Bibiiography ) .

On the other hend- there have beensuggestions that a porous crust might harbournoisture, or alkaline solutionsr or evennicro-organisms ancl thus enhance furtherr,reatheri ng (net . f Z5 ) .

Yet again' Dr R.H. Brifl (Ref . fl+) hassuggested. that argr vigorous cleaning of theglais will remove the protective fire-finished surface (vhich is known to bespeciaffy durable in nevly-mad.e articles) ancl

trence vi1l enhance further veathering'

Moreover Klaus Ktlhne (Ref. B)+) trasclained that polishing of the surface vithacicl improves the d.uratitity of the glassbecause it recluces the alkali eontent of theglass surface (fire-finishing has the sanemeehar:.ism, the atkali being reclucect byvolatilisation in the f1a:ne).

Tllus the problem clearly neecls someexperimental stucLy but there are nany diffi-culties in carrying out suitable experiments,espeeially as ac-celeratecl veathering must beused to obtain answers in a reasonabl-e timeand. the acceleration proceclure may introd.ucesome d.istortion in the a.nsvers. Howevero astart must be made someuhere, somehov, anclthe folloving brief notes of tvo erperimentsbeing caried out at BGIRA in Sheffield. areincluded in the hope that they may encouragesomeone 1s ssmqnt and perhaps inspire some-body to carry out a better experiment!

Britr (nef. l-l+) ctaimed that loss of afire-finished surface vould. enhance weather-ing anci hence the first experiment used apoorly durable simufated as4lslra."l glass(sritish simulated glass No.2) having afreshly-neltecl fire-finished surface. A l7nmd,iameter disc was cut from the cast slab ofglass and a narrov stripe vas airbradecl acrossthe flat surface. After exposure for 5O hoursto disti[ecL water in a Soxhfet apparatus theairbraciecl stripe hacl turnecl a cleep red- becausethe iron in the clark green glass had beenattacked. where the durability of the glass had.been lessened. by the airbrasive treatment.fhe rest of the surface was still green andshiny but the sicles of the dise, where it hadbeen cut from the slab of glass vith a diamonclcore-dri11, had also turned reil.

A seconcl stripe vas then airbratled atrigbt angles to the first one ancl the sarnplevas again exposecl for 50 hours in the Soxhlet.The red cteposit vas lost frorn the first stripebut another developecl on the secontl stripe.These effects can be seen in Figs. I a^nd- 2.Fig.l shows the d.isc and the retl cieposit onthe second stripe shovs up as an irregularvhite rnark (presunably the airbrasive gritblev across the .lisc towarcis the l-eft anclproduced the ragged edge). Fig.2 shovs anenlarged view of the central portion takenvith extreme oblique illr:mination and the

position of the originet stripe can be seenas a groove vith a shaclov in it.

" Thtrs-r&he airbrasive untloubtedly reduces

the clurability of a freshly-melted. fire-finished surface, but the process of acceler-ated. veathering seens to confer fresh clura-bility on the initial stripe beeause it tloesnot fo:m a seconal red. ti,eposit. Moreover, ved.o not knov hov J-ong a fire-fi.nished. surfaceean remain as a protection against corrosionwhen exposed to the weather. In orcler toarsver this seconcl question (at least in part)the next experiment usetl a smal1 piece of12th century gJ-a.ss from York Minster, by kindperrission of the very helpful Dean andChapter. Mr Peter Gibson carefuily selecteda piece of glass having the follovingcharacteristics: (i) it should. be as similaras possible to the piece of 12th Centurygreen glass clescribed. in Ref, I29 and hence]-ikeJ-y to have a poor durability, (ii) itshoulcl be completely covered in a eoherentcrust, including the edges, (iii) there shoultl-be as 1itt1e paint on the surface as possible,and. it should shov no evicl-ence of ever havingbeen broken (i.e. apart from grozing it shouJ-dhave the original surface), and (iv) it shouldbe sma1l enough to fit inside the Soxhletapparatus; the actual climensions vere25x18x3nm.

A piece vas found vhich met theseexacting requirements and- it veigbed. 36ZZ ng;its volume was about 1.\ nl- ancl its surfacearea wa.s about 12OO rnn2. The first e:q>eri-nent (a) vas to vash it well in orcler toextract any alkali (potash) r^rhich may havecollectecl in the crustl this was clone byvashing for 30 ninutes in hot distilled vaterat 6ooC. A substa.ntial a.mount (0.73 ng) ofpotash (f2O) vas vashecl out from the crrrst,representing O.O2o% of the veight of thesample (the period of immersion was too shortto perait any attack on the gJ-ass). text (U)the crusted sample vas put in the Soxhletapparatus for lO hours at about BSoC and itvas found that a further 0.78 nrg of K2O wasextracted (O.OZZ%) or about 0.5 ue K2o per nm2of surfacb in 5Oh. Some of the crust beca,mecletachecl cluring this prolongeil exbraction, sothat the nev rreight of the sainple vas 3555 ng(thus about 5li ng of crrrst was probably lost).

Next (c) the crust on both sides antl allfour ed-ges vas cJ-eaned of,f vith the airbrasive(using No.3 erit) anal the cleanedl sampleweighed J20O ng. ft is nob known horr muchglass vas removecl cluring the cleaning processbut the loss in veiglrt ot 356 mg (about 10%)probably represents mostly crust. The cleanedsample vas then put in the Soxhlet for !O hoursand 0.55 ng of K:O vas found in solution inthe extracting vater (O.OZL% of the new sa,mpleveight). The sample lost a further l+.1+ mg inveight which possibly representecl loss ofaclherent alumina grit fron the airbrasive a^nd.

perhaps other losses of glass, erust, etc.Up to this point the apparent durability ofthe glass could be rep3esenteil by a J-oss ofabout 0.022% ot its weight as K2O in 50 hoursat 85og vhether the crust rrras present or not.

Fig. 1 fhe circle is a 17 mm diameter disc of poorly'durable darkgreen glass with a fire-finished surface which has had a

stripe airbraded across the vertical diagonal. After accelerated

weathering, the iron in the glass weathered to give a red

deposit which appears white in the photograph. lt was

therefore concluded that airbrasion of a fire finished surface

reduies the durability of the glass.

The letter A indicates the sane spot in bothFigures.

Thus it seems from (a) ttiat the crust(which a!-sos.!- weighed l+ZO ng) contained.0.73 ne of QO or O.Il/0, and this may 1-re

evidence of trre erust Itharbouringtt alkal-inesolutions. Moreover, fron tc) it seens that

*".dleaning a'ith the airbrasive did nol changethe apparent durability of the glass.

uexU (d) the sa:npIe r.ras grounci smoothusing Carborunclum grades 22O, 32O and 5OO,

anal then polished mechanica"lly usingttCerirougett 688, Grade 90. This reduced theveight consiclerably (to fToB ng) and the neqd.i.nensions uere 2I+ x 15 x 2 mm, or aboutO.T uJ- o rdth a surface €Lrea of about 950 nn2.It was again given a Soxhlet extracf,icn fo!'!O hours but only 0.21 ng of K20 was extractedio.orz of the nev veight or o,V ug K2o per m2of surface). The conclusion is that theapparent durability of the suriace has beengreatly increased, being cloublecl orl the basisof percentage loss of K2O (on1y ha'l,f as much

lost) or increased. about 3 times on the basisof loss of K20 per unit area of surface.

From this it could appear that the crustpresent at stage (l) aia not have a protect-ive action a"nd that polishing of the glass(stage cl) enhanced the clurabilityt probably byreclucing the total effective surface roughnessancl possibly also by bringing about some lossof afkali from the heating of the surfaceduring wet grinding.

TLre ex;reriment is not yet completed'beeause stage (e) is involved vith theproiltrction of a flame-finished surfaee, andthe res'*fts .,ri11 be reported. in the next Nevs

Letter. (Tfre first part of this study issupported by the Royal Society and the secondprtt ly the Department of the Environment.)

Ttre results to d.ate are srunmarised in thetable on page , for convenience.

A

Fig. 2 lhe stripe in Fig. 1 was actually the second airbraded line to be applied and the first can now be seen as a trough in oblique illumination.

It also went red after the first period of weathering but it did not go red again during the second period of weathering. lt has therefore been

concluded that further weathering of an airbraded surface may restore some of the durability.

Procedure Weight ofsa'nple (ng)

Percentageloss of K20

Loss of I!20per unit area

Coments

(a) ilashing

(l) Ertractionthrough thecrust

(c) nxtractionafter airbrasion

(d) nxtraetionafter polishing

3 HEATTNG OF MEDIEVAL STAINED GLASS !

I feel that the time has come to re-sxaning the valid.ity of the videly-heId. viewthat most medieval stained- glass should not bere-heate{l at all. Here I includ.e the possi-bifity of tvo types of re-heating, thetrsubstantialft heating (e.g. to 2o0oC) vhichnight be involved in ttbakingtt a resin on tothe surface anCI the [extremett heatinginvolved in refiring the g1ass, either tofire-on reneved, painted lines or to proilucean rena,melletl.tt (or even a flarre-fired.) rinisrras a means of inproving the clurability of theglass.

Probably the first paper aClvising againstthe heating of renaissance glass vas that byMiss Mavis gins6G-dl? Tony Werner in t95l+(see Ref. 9 of the British Acadentrr Bib1io-graphy). They founcl that the surface of alJth Century goblet became badly crazed byheating sIowly to on\y gOoC over a period ofB hours. As a footnote to the sa,Ine paperDr R.H. Bril1 cornmenteii that, at the CorningMuser:m of G1ass, 23 sanples of rrancientglassesrt hatl been heatetl (first to I10oC for60 hours ancl then to 1?5oc for 160 hours ) antl30% ot them vere clamageCt.

These experiences vith renaissance vesselglasses tay not be vho1ly relevant to thE-question of heating ear\r neilieval wind.or,rg1ass, if only because the nedieval vinclowshave been durable enough to renain extantTffib"eh d"e"ged) for 5oo-Boo years and. itvoulil be i-nteresting to speculate on the fateof renaissance goblets if they were to beextrlosed continuously to the veather for asinilar period!

Hovever, vhat ve rea11y want to knov iswhat aetually happens to metlieva1 wind.ov glasswhen it is heatecl. In 19?1 Dr Gottfried.Frenzel tolil ne that there was rarely eny

i ue/o*2 )

(J.b Alkal-i hefd. in thecrust

The crust is notprrctective against thistype of veathering(5hng crust atso lost)

The airbrasivetreatment d.oes notreduce the d.urabilityof an ancient sample

TLre resista,nce to thistype of weathering isat least doubled bypolishing the sa.rnple

o.2

clanger in heating medieval- vindor.r' glass tottextremett temperatures provitling the surfaced.id. not alreacly shov signs of ttmicro-cracksft.In fact, much medieval painted. glass washeated- to 5OOoC to 5OooC when the itsuperf\rsionprocesstt (see Dr Eva Frocl-l--Kraf't, Ref. l+l+(b)tfor tletails ) r'as in vogue for conservingstaineil glass ancl it ca,me to no harm. In theSt Lorenzkirche in Nuremberg 60 panes r,rereheated to at least l+OOoC in 1919 and another90 vere so treated. in 1938 without harm (seeRef. 35). Josef Schnitz repainteil and refiredglass in the St Sebaldkirche (OieDenbalpflege, 1919 2]-, 97-99, tO5-rOT) andRud.olf Pfister did the sane in the St Lorenz-kirche (Deutsche Denkna.lpflege, f939 66-tB).

In lten 1-A of the Bibliography prepareclfor the 1972 CVMA Colloquium, anci- datecl 21August 1972, I enquired vhether erqr practisingrestorers or any tlelegates to the ColloquiumcouJ-d provide arly evidence of med.ieval glassbeing d.a.maged by heating. Onfy Mr DennisKing of Norwich sent me a reply antl he statedthat there vas generally no risk, but a^rqr

glass vhich shovecl surface fracturing (suchas that fron St Michaelrs Catheilral, Coventry- see Fig.3) vould be liable to d-a,nage byeven moderate heating.

Modern atlhesives usecl for edge-joiningbroken pieces tencl to need some heat if theyare to ttcurett reasonably quickly ancl vetherefore need to know vhether there is anyrisk of heating meclieva.l r.rindor^r g1ass,provicling it d.oes not alreaftr shov signs ofcracking such as that in Fig.l+, and- proviclingthe surface has been carefully cleaned beforeheating (see Ref. U+).

I therefore reIgE_.SLjpEf.; doesar\yone knov of caF@fffi has beencla.rnagecl? Please vrite to me at ) HardvickCrescent, Sheffiefd S11 BvrB.

3622

3555

3200

1708

0.020

o,o22

0.021

0 .012

/

Fig.3 This glass from St. Michael's Cathedral Coventry shows

surface fractures which cause concave cavities to form; itwould be damaged by heating.

Fig.4 This glass shows surface fractures which link up and would penetrate the glass if it were to be heated.

4 PROTECTIVE GLAZINGS

)+.T ISOTHIRMAL GLAZING

Nevs Letter No.1O containecl, a.s item 3.1(a.nd. as Fig.l+ of that issue)r some tletaifs ofthe isothermal glazing installation at theChurch of Santa Croce in Florence, but no

measurements vere avail-abLe at that tine.Professor Marchini has nov kindly sent me

some measurements vhich are shown d.iagramat-ical-ly in Fig.5. They can easi\r be relatedto tffelfrotograph in N.L. l{o.10. The centralventifator wilh three apertures has a totalarea of 2'l cma a^nd the tvo side ventilators,vitn tivq apertures, each have a total areaof l+5 cnz. fhe piece of glass vhich formsthe external glazing to this tracery isremarkably l-arge, being 2.1!n x 1.T?nl

Higo-i

mftl

f4d,i<,ltal 6lass

?'l5Onr -

SEcrro,tA-A

jll ht

lvtod.evr.. 6to.ss

Frg.5 Details of the isothermal glazing system on the tracery lights at Santa Croce Church, Florence.

It,z FXTERNAI PROTECTIVE GLAZ,ING AT YORKMINSTIN

By kind" peruission of the Dean and-Chapter, studies are being carried out of thechanges in hunidity insid.e the space betveenthe med.ieva-l glass a.nd. the external- mod.ern

Date Time

dianond-quarry glazing of tvo r,rind.ovs at YorkMiniter, a hair hygrrcmeter having temporarilybeen inserted into the space.

Great East Wind.ov and the folloving data vereobtained:-

Weather conditions Insid"e Outsicletemp. temp.(oc) (oc)

RH

(%)

Moisturecontent

r r ?r(g/n-/

1l+ 0ctober

]5 0ctober

l-6 October

17 October

18 October

l-9 October

20 October

21 October

oB. r-o10.\518.35

u6.L)4r_0 .0018.36

OB. OB

10.2018.l_0

08.03Lo.25l-B .20

oB. o5r0. l+0

r8.25

08.1510. \0IT.2O

09.2OL..+)18. 30

oB .05

Dry, sunliglrt on vindovDry, sunny*Dry, clear sky

Cloudy, light rainCloudyHeavy cloud

Cloudy, light rainClou{y, no sunHeavy cloud

OvercastCloudy, intern,ittent rainModerate rain

Heavy rainOvercast, ligirt rainLight rain

Dry, sun on vind.ovDry, sunny intervalsxLight rain

Dry, veak sunlight on vindo.ntDry, cloudy, no sunDry, cloudy

Dry, cloudy, no sun

x But the sun was not shining on

B.g9.)+8.9

O-U.J8.98.9

8.38.98.9

8.98.38.9

9.48.38.3

8.38.9O.O.J

8.98.98.9

oOl.o

the glass

{.oro.6no

aa

o.JnOl.u

o.r9 .1+

T.B

6.tfI.77,2

o.fL2. B

10.5

'7D

10.08.9

o.ll-0.08.9

5.0

$.e)( 5. )+)

(6.5)

7.9(.9rya

'70eal.oT .l+

(8.:)AE

t.)

(B.e)l.J7.1

(>.2)(5.2)6.8

1-u. {

6.6o.l

(r.:)

ol7075

959LB9

9l+B9B:

9B90B6

9BB9B6

5I+

7IB1

T775T7

9r

The ternperature insid.e the Minster vasmeasured. vith a thermometer close to thestained" glass at the position of the hygro-meter; there vas littJ-e variation, betveenT.Bo and 9.\oC. The outsid^e temperaturesvere measured at the back of a nearby housein a comparatively shel-tered" position and thevariation vas betveen 50 and. 12.BoC. Themoisture contents of the air in the inter-space vere estimated. on the assumption thatthe temperature in the interspace vas that ofthe Minster; this coul-cl lead to the errorsvhich are d,iscussed. belov. Ttre relativehunidity in the ear\r morning tendecl to bein the range )l+% to 9B%, unless the su:: vasshining on the vind.ov, vhen the va] ue vas6\% to l7%. However, the figures for themoisture content of the air are more inform-

ative. Except for the eight fi.gures 1nbrackets the v4lues l-ie betveen ?.1+ and ?.9 gof vater p"" 13 in the first part of theperiod but seen to fall to 6.6 to l.l g/n3in the second" part; the corresponding dev-points would have been 7.5 to 5.5oC or )+.5to 6.2oC. The five lov vatues for l-l+th andIlth October (especial]-y those of 5.9 and5.2 g/n3) occur vhen the sun vas shining onthe vindov or the sky vas brieht; in thesecircr:mstances the temperature of the air inthe interspace vou-l-c[ certainly have beenhigher than that record-ecl for insicie theMinster. In fact, a moisture content of7 .9 e/n5 and a relative hr:midity of 67% \rouldcorrespond. to en interspace temperature ofabout 11.!oC vhich is only about )+ degC abovethe temperature in the Minster.

SimiJ-arly, the three "hightt values of8.5, 8.9 and ?.3 g/m5 correspond. to rathercold nights and the interspaee temperatureon 17th October correspond.ing to 'l $ glm3wou1d. have been about 7;oC, betveen the vaLuesrecord.eii for ttinsiclett and. troutsidett. lt seemspossibl-e that some cond.ensation may haveoccurred on ttface 2tt (the insiile of the outerglazing) on the nights at L5/I6t]n and.zo/zl.st'October. It seens likety that the moisturecontent of the interspace vas actually loverin the seeonii part than in the first part and,if Lhis vas so, it night comespond vith thefact that the veather vas rainy until- l8thOctober and d.ry thereafter. This voulcl

5 NEW ABSTRACTS

a75. CARLSON, W. (P>e) rtlnterfero-metric stud-ies of convective fLov phenomena invertical, plane enclosed air layersrr. PhDThesis, Universi.ty of l{innesota, April 1p!6.298 + xvii pages of typescript.

This long thesis is highl-y technical- butits conclusi.ons can be of interest to archi-tects and others concerned- wi.th the technicalrequirements of isothennal glazing.

The author determined- the directions andextent of air flovs i.n a tall narrov cavity-rrhen one sid.e is hotter than the other. Manyof his results are not appropriate for thesituation of cathed.ral- vindovs (e.g. thecavity vas only B mr wid.e and- the temperatured.ifference vas 83 deg0) but some of his testsrrere made vith a cavity I52 nm vid"e and atemperature d-ifference of B d-egC, so these areof greater interest.

At the narrovest spacing (B nm) the mainheat trarrsfer va.s by conduction and. thetemperature profiles across the interspacevere linear, but as the interspace vas mad.e

r^ri-d.er the mechanism of convection came intoeffect. Air floved up the hotter surface andd-ovn the colder surface vith horizontal flovsat the top and bottom, and this beca,ne notice-abl-e vith the B nn spacing.

End effects rrere founC at the top andbotton of the cavity alld both of these effectsbeca.me more pronounced vhen the temperatured,ifference exceed.ed 20 degO.

Even at the rridest interspace used.(flZ nn) the central core of the air layer hada temperature grad.ient, being varmer at thetop than the botton. In these cases the flovbecaJte turbul-ent.

suggest that there is some ttleakagett of airfrom the atmosphere into the interspace.The clata also inalicates that there is notrreservoirff of cond.ensed. vater (or rainvater)in the interspace or it vould have increasecithd"Iiftidity on those aays (fl+tfr and. 19thOctober) vhen the sun lras shining ancl theinterspace vas \rafitr; this confir:nrs thesuggestion that the interspace is ttl-eakytt.

Further stuclies (to Ue reported. in asubsequent Nevs Letter) wi11 be carried outon a south-facing windov (Uo.l?, in the SouthChoir Aisfe, seconcl from the vest).

W. oLBRrcH, H. (rsr:) "Molasses as

remedy at the renovation of the glasspainting of a church vindov of St Lorenz inNurembergrt, F.O. Lichtrs International-erMel-assebericht. International MolassesReport, 10 (6) ppl-3 (12 April 19?3).abstract-is taken in its entirety fromVoI .11 No.1, Sr:rmer 19?l+ p.T5 (abstract11-2?3) and- there is as yet no furtheration.

Thi sAATA

inforn-

The use of molasses in the conservationof a painted glass vind-ov (the Hal-lerfenster,mad.e in t-l+BO, Uy the Micha.el Wolgemut Work-shop) is discussed. In 1939' as a temporarymeasure, broken fragnents were joinetl usingan adhesive made from nolasses and a d-extrin-protein g1ue, sntr the repairedi pieces vereprotected on either sid.e by tvo sheets ofclear 'glass. The treatlrent is stifl consitler-ecl- satisfactory after thirty years.

]JT. PATERSON, M.P. (197\) "Ttre atmos-phere of the Royal A.l-bert Hal-l - a stuQr ofthe internal atmosphere of a natura-l- tlrauglrtbuil-dingrt. Air Pollution Research Group,Matheuratics Departinent, Tmperial CollegenLondon S.W.7. 12 pages of typescript, 28thOctober 1971+.

The atmosphere of the Roya1 Al-bert HaIlin London is of interest to architects con-certred. vith cathed-rals because it is a largebuilding (:-05*3) which can holcl a large nurberof people ( about BOOO ) and its atnosphere canbe controlled by only tvo adjustable factorsl(a) by opening outside doors and vindowsn etc.and (l) by turning on the heating system.There are no fans for forcing air through thebuilding and- no cooling systeu.

o

Ttre height of the building (:O r) assistsmovement of the air by thermal- convection(prod,ucing a pressure difference in the coronaof about 0.02 nb per degO of temperatured,ifference) anct the thick va1ls prcduee athermal inertia; vhen there rrere no occuparrtsthe inside temperature varied by 1 degC vhilethe outside temperature charrged by 9 degC.The study vas mad"e during the BBC PromenadeConcert Season and. an ar.ld.ience of 5000 personsand 100 perforarers vere assumed to contribute6ao tI'I of heat and \l-2 kg/n of water vapour.There vould be an average ot )+50 kW fron theheating system during the vinter (f5OO t<w atpeak rates) a.nd 12@ trg/h of vater vapourbrought in by the florr of outside air (assrm-ine 60% relative hr:nidity' 22oC and thresidence time); the lighting cluring aperformance contributes an extra 95 kW.

During an average perfo:mance of 2l hoursthe temperature rises by about 2.5 deg0 (from22.O to 2\.5og1 anci moisture content byz.B g/n3 (the rel-ative humidity from 60.0% to6l+.8%). (RGN - he calculated. that the mois-ture content night rise by l+.t g/n3, and thetrue val-ue probab\r lies betveen these trrofigures because hair hygrometers tend to givel-orr va.Iues vhen the hr:midity is risingrapia:-y. )

(Comment by RGN - We can draw sonei.nteresting deductions fronr this paper aboutthe vay in vhich a J-arge congregation vouldhelp to cause condensation on the inside ofthe vindovs of a catheclral. In London theaverage outsid-e d-ev-point is 3oC, corres-ponding to 6 g/m3 of moisture in the air.The congregation vould add 2.8 g/n3 and ttreresulting moister air (B.B g/n3) vould have adev-point of !oC; if ve assume that theinside of the rrindov is at the same terper-

ature as the outsid.e air, conclensation vould-occur 90% ot the tine in an average January.fn the stumer, the average d.ev-point rsould. beraised from 12oC to 15oC and., again naking thesa.ne assumption, cond.ensation vould occvr \O%

:o'? Ttre tine in an average July. )

Dr Paterson tells me that he vou1d. beglad- to hear from any read.ers of the NeirsLetter vho have cornments on the situation.

1?B. RYND. J.P. and RASTOGf r A.K. (19?l+)"Auger electron spectroscopy - a nev tool inthe characterisation of glass fibre surfacesrr.Cera,n. Bul-l., r97\ 53 6Sf-6:l+, 5:7.

fhis is a rather technical articl-e aboutan expensive piece of equipnent vhj.ch usesel-ectrons to stu$r the surface composition ofsa.mples of glass. It is of potential- interestin the scientific study of corrosion phenomenabecause it stud-ies only the layer of atoms atthe surface. The authors found that thecompositions of the surface layers of tvonevly-formed. moclern glasses (glass filrecompositions E a^nd S) tiffered. from those ofthe bulk glass. The surfaces vere, forexa$ple, found. to be richer in some el-ementsthan in others but only al-r:minium (al-urina)vas richer at the surface in both glasses.

o-o-o-o-o-

NOTE: Will reaclers of these Nevs Lettersplease drav my attention to any papers vhichshoul-ct be abstraeted here. It rrould beparticularJ-y helpful if photocopies of thepapers cou.J-d. be supplied. My acld.ress is!, Harilvick Crescent, Sheffield,, S11 BWB,England.

l,II

NOTES

I

I

10

Reproduced by the University of York Printing Unit