XIX IMEKO World Congress Fundaental and Appled Metrology Septeber 6, 009, sbon, Portugal IMPROVEMENT OF THE REAIZATION OF FORCES BETWEEN MN AND 5 MN AT PTB THE NEW 5 MN FORCE STANDARD MACHINE Falk Tegteer, Rolf Kue, Mark Sedel Physkalsch-Technsche Bundesanstalt, Braunschweg, Gerany, Falk.Tegteer@PTB.de Physkalsch-Technsche Bundesanstalt, Braunschweg, Gerany, Rolf.Kue@PTB.de Physkalsch-Technsche Bundesanstalt, Braunschweg, Gerany, Mark.Sedel@PTB.de Abstract Snce Noveber 008, PTB's force scale has been copleented n the range fro to 5 MN by a further force standard achne. Ths autoatcally workng 5 MN Force Standard Machne ( utlzes the hydraulc aplfcaton of a 50 kn ass stack and enables low uncertantes of saller than 0.0% by usng nnovatve ethods for the control prncple and the lnk-up of the force standard. In the paper, the constructonal desgn of the achne, the control and the nnovatve lnk-up procedure wll be publshed. Suppleentary to ths, results fro coparson easureents of the achne wth PTB's and 6.5 MN are presented. Keywords: Force standard, hydraulc aplfcaton, force transducer. INTRODUCTION In the past, PTB only had the 6.5 MN force standard achne ( at ts dsposal for the nvestgaton of force transducers wth easureent ranges larger than MN. Due to the great deand for calbratons n ths upper force range, ths caused, te and agan, bottlenecks as ths achne was not always able to satsfy the great deand for easureents appled for. Aong the great nuber of orders, the share of transducers up to 5 MN aounted to approx. 80%. For ths reason, PTB has decded to ake use of a hydraulc force standard achne taken over fro the forer ASMW ( Offce for Standardzaton, Metrology and Coodtes Testng n the forer GDR and to utlze t - after ts coplete odernzaton as force standard achne - for forces up to 5 MN. It was aed at reducng the uncertanty n the force realzaton of ths achne less to 0.0%. At the sae te, the updated faclty shall allow an effcent, autoated operaton. Innovatve solutons for the control and for the ethod used to lnk up the achne wth the standards have been nvestgated and realzed.. CONSTRUCTION OUTINE OF THE MACHINE The renewed easureent devce (5 MN force standard achne, Fg. works n accordance wth the hydraulc aplfcaton prncple. Ths eans that - frst of all - the weght forces of a 50 kn ass stack (step sze 500 N act on Workng cylnder Cross head as support for the - two workng cylnders and the upper pressure plate Force transducer for copresson Cross head lfted by the two workng cylnders Mass stack and easurng cylnder Force transducer for tenson Fg.. 5 MN Force Standard Machne.
a pston-cylnder syste on the devce's easurng sde. By a control procedure, whch s new for ths achne type, force equlbru s establshed between the weght force of the weghts and the hydraulcally generated force. The control procedure coprses the cascade connecton of a control unt for hydraulc pressure, resdual force and poston. Due to the constructon, the achne s a hghly nstable syste - n favour of saller echancal uncertantes n the force realzaton. A sophstcated, coputer-aded control unt allows the equlbru of the forces fro the ol pressure on the easurng cylnder and the ass stack to be establshed wth uncertantes saller than 0-5. The ol pressure requred for ths force equlbru acts sultaneously on two pston-cylnder systes arranged n parallel on the operatng sde. Due to the relaton of the surfaces of the pston-cylnder systes on the operatng sde and on the easurng sde, the forces are hydraulcally aplfed by a factor of approx. 00 whle the pressure reans the sae. Both pston-cylnder systes exhbt a coparably large gap whch - even when hydraulc ol of vscosty grade 00 /s at typcal achne teperature s used - causes a consderable loss of leak ol (7 l/n. To rule out that the leak ol flow ncreases too strongly when the pressure s ncreasng, a syste wth a double wall s used to prevent that an elastc enlargeent of the cylnder occurs. At the sae te, soe ol nlets were provded whch ensure a stable gudng of the pston thanks to a pressure whch s unfor over the entre crcuference. In ths way - wth the larger gap and the ol nlets - t can be avoded that the pstons touch the walls of the cylnder - whch would lead to hysteress-affected frcton. Into the two pstons theselves, one hole each has been drlled at two thrds of the heght, and the pstons transt the force fro the botto of the hole towards the upper crosshead va pressure bars whch are sphercal at the contact surface. Ths desgn prevents that torques are appled to the pston whch could otherwse press t aganst the cylnder wall.. INKUP OF THE MASS AND UNCERTAINTY- MODE When a force s generated by hydraulc transsson, a force of gravty n the gravtatonal feld of the Earth s - n analogy to drect loadng - frst actng on the weghts. Ths force s then transtted by coupled hydraulc pstoncylnder systes. The force generated by the hydraulc transsson prncple s descrbed by the followng odel functon: F = g ρ ( Q ρ The nput quanttes are: ass of the weghts g al gravtatonal acceleraton at the place place where the weghts are nstalled ρ densty of the weghts used ρ densty of the ar Q transsson rato = ( ( enlargeent of the pston-cylnder systes by ol pressure force ntroducton effects for an deal test pece nfluences by agnetc propertes of the weghts In the past, an uncertanty odel for coparable hydraulc standard easurng devces was establshed n accordance wth ths odel functon. It turned out that the uncertantes n the deternaton of the transsson rato and the exact weghng of the asses ake a consderable contrbuton to the resultng cobned standard uncertanty. Ths s why the 5 MN force standard achne shall be lnked up va a dfferent ethod. The asses and the transsson rato are not prarly deterned for the uncertanty budget. Due to the place where the 5 MN deadweght force standard achne s nstalled -.e. besde the MN deadweght force standard achne - lnk-up wth ths devce was perfored by transfer transducers. The dentcal envronental condtons wthout sgnfcant transport nfluences and delays n te worth entonng allow the devce to be lnked up wth saller uncertantes than has before been possble by the easureent of the pston-cylnder systes and weghng of the ndvdual asses. In the case of the 5 MN force standard achne, the change of the transsson rato alone aounts to approx.. 0-4. The 5 MN force standard achne was lnked up wth a buld-up syste coposed of a transducer for forces n tenson and forces n copresson ounted n parallel as well as of dfferent transfer standards for forces n tenson and forces n copresson. Equaton descrbes the uncertanty odel for a hydraulc force standard devce n accordance wth the odel functon (. w( F = w ( + w ( g + w ( ρ ρ w ρ ( ρ + w ( Q + w ( ρ ρ = + ( Ths odel shall not, however, be used for lnk-up of the 5 MN force standard devce, as t would not allow the absolute values of the ndvdual uncertanty contrbutons to be deterned wth uncertantes as sall as they are acheved by drect lnk-up wth the MN force standard devce whch s explaned n the followng. In [], the best easureent capablty s descrbed for the lnk-up of force easurng devces by eans of transfer standards. Chapter 5. relates to the lnk-up of a hydraulc force easurng devce. The followng uncertanty functon s obtaned:
w( F = w ( + w ( g + w ( ρ ρ w ρ ( ρ + w ( Q ρ ρ + w ( Traceablty Sawla [] defnes the ter of the lnk-up w ( Traceablty uncertanty as follows: w( w ( F Traceabl ty = w ( F + w ( RelDev + w ( + w ( Hys Realzato n + w ( + Drft_TtaStd ( + (4 easureents and ass correctons, thus corresponds to the uncertanty of the transfer process. The followng s thus vald: w ( + w ( g + w ( ρ + w ( ρ + w ( Q = w ( Hyst. Tr.wthout ρ ρ ρ ρ As the asses were always copared wth seres at ncreasng forces, the hysteress or reversal error s of no sgnfcance for the lnk-up. If equaton s copleented wth equaton 5, the followng uncertanty odel s obtaned for the 5 MN force standard devce: (5 Where: w( F = w ( Tr. wthout Hyst. + w ( Traceablty (6 w ( F = n = ( x w ( = x n( n x a Hys Hys a Drft TraStd Drft_TtaStd w ( = w ( F = n = ( x x n( n x ( x x w ( RelDev = = 4 x areldev w ( of force 5 realzaton = 0 6 (Standard devaton n the realsaton Ths lnk-up uncertanty s now used n two ways for the uncertanty odel of the 5 MN force standard devce. The uncertanty odel on whch equaton ( s based contans agan - as n equaton ( - the uncertanty contrbutons of the force coponent generated on the sde of the easurng cylnders as well as the uncertanty contrbutons of the transsson rato. In the case of the 5 MN force standard devce, the assocated quanttes were - as descrbed before - not traced back by ther explct deternaton, but by a coparson between the force obtaned as overall result and the reference force of the MN force standard devce. Equaton 5 now adds up the achne-relevant coponents fro equaton. The uncertanty contrbutons of the sngle quanttes were not deterned drectly, but the resultng cobned uncertanty was quantfed wth the force coparson. The etrologcal uncertanty wth whch the asses were algned as exactly as possble by coparable For the ass lnk-up of the achne, the asses were weghed. Subsequently, coparson easureents were perfored wth the or 6.5 MN-force standard devce. On the bass of the devatons, correcton values were calculated and the tare weghts were adapted untl the devatons becae, f possble, saller than. 0-5 n the range fro 00 kn to MN and saller than 4. 0-5 n the range of the easureent wth the buld-up syste coposed of the MN transducer for forces n tenson and the 5 MN transducer for forces n copresson. Mechancal superstructures for such a buld-up force transducer arrangeent - e.g. a large copresson plate under whch several transducers are ounted are, for dfferent reasons whch are related wth addtonal echancal coponents - not precse enough for the lnk-up of the achne. Ths s why the possblty realzed on the 5 MN force standard achne -.e. drect, sultaneous nstallaton of both a transducer for forces n tenson and a transducer for forces n copresson - s ade use of wthout causng addtonal uncertantes by addtonally ounted parts. In the dfferent ountng postons and wth rotatonal turnng of the transducers, uncertanty coponents below. 0-5 could be detected due to the rotaton effect. The transducers had before been calbrated n the MN force standard achne. For ths purpose, a force-n-copresson transducer up to MN was used, whose nterpolaton functon had been deterned n the MN force standard achne and confred n the 6.5 MN force standard achne. Fro the addton of the easureent values of the transducer for forces n copresson ( MN and the transducer for forces n tenson ( MN, the achne was lnked up n the range fro to 5 MN to the hgh-precson MN force standard achne. The photo n Fg. shows both transducers ounted n the 5 MN. Fg. shows the sgnals of both transducers durng a easureent fro MN to 4 MN. The saller transducer for tenson wth ts long bars for ountng s ore elastc than the bgger transducer for copresson. The exact rato of force between the two transducers s fnally adjusted wth the screwed head of the copresson transducer.
Dsplayed force n kn 500 000 500 000 500 0 Couplng pont of the stffer copresson transducer Tenson transducer Copresson transducer Coparrson area 0 500 000 500 000 500 000 500 4000 oad step n kn Fg.. Caracterstcs of the two transducers n the buld up syste. The tenson-copresson tests were perfored n two ranges: n load steps fro 000 to 4000 kn and n the range fro 4000 to 5000 kn. In the last-entoned range, the transducer for forces n copresson up to MN was used. It showed, however, a clearly stronger creep behavour than n the range up to 5 MN. Accordngly, t was calbrated n the 6.5 MN force standard achne. In the calbratons on the MN force standard achne and on the 6.5 MN force standard achne, the force steps and te dependences were adapted exactly to the force/te curves of the actual tenson-copresson test. The MN transducer for forces n tenson was, for exaple, calbrated n 0 steps of 500 to 000 kn to acheve that the force curve corresponds to the curve shown n the dagra n Fg.. Fro these calbratons, the nverted rd order polynoals to the force-sgnal curve were calculated and used for evaluaton of the ndvdual easurng sgnals of the two transducers. Table shows a lst of the ndvdual results of a easureent, ther addton and devaton fro the theoretcal value. Table. Sgnals and force value durng the calbraton as n Fg.. Force Sgnal Sgnal Force Pressure Force Tenson Addton rel. Devaton V/V kn kn 000,597 0,8548 568,444 4,60 000,054 0,007% 00,6858 0,5898 608,99 59,046 00,09 0,008% 400,67878 0,86 648,04 75,706 400,0 0,0004% 600,7805 0,906 686,70 9,0 600,00 0,000% 800,7566 0,4684 74,96 075,66 800,0 0,0008% 000,7946 0,5684 76,550 8,480 000,00 0,000% 00,85 0,6076 798, 40,86 00,048 0,005% 400,8698 0,67880 84,496 565,57 400,0 0,0007% 600,90607 0,74 870,44 79,48 599,94-0,00% 800,945 0,8575 906,00 89,97 799,957-0,00% 4000,97876 0,8806 94,445 058,50 999,965-0,0009% After the easureents had been adjusted as exactly as possble, a concludng seres of coparson easureents was perfored. The ndvdual paraeters for equaton (4 were also deterned fro the concludng coparson easureents. The results are shown n Table. The Table also contans the values for the two Traceablty factors and the sple easureent uncertanty of the 5 MN force easurng devce calculated n accordance wth equaton 6. Table. stng of all uncertanty coponents and the resultng cobned uncertanty. Force n kn Reference x x w rel devaton w w Hys w w D realzaton w t traceablty w tr wthout Hyst. w (F 50 -MN- 0,0005 0,000,50E-0,07E- 7,6E-0,46E-0,00E-0,68E-05,50E-05 4,45E-05 00 -MN- 0,99994 0,0000,9E- 5,60E- 7,6E-0,77E-,00E-0,98E-05,7E-05,4E-05 50 -MN- 0,99968 0,99978 4,6E- 7,E- 7,6E-0 8,57E-,00E-0,98E-05,7E-05,4E-05 00 -MN- 0,9996 0,9996,66E-4 8,47E- 7,6E-0,77E-,00E-0,94E-05,7E-05,6E-05 400 -MN- 0,799888 0,799898 5,7E- 5,E- 7,6E-0,5E-,00E-0,94E-05,7E-05,6E-05 600 -MN-,9985,9984,8E-5 5,4E- 7,6E-0,9E-,00E-0,89E-05,04E-05,07E-05 800 -MN-,59970,59979,5E-,4E- 7,6E-0 7,6E-,00E-0,89E-05,05E-05,07E-05 000 -MN-,999598,999605 6,06E-,99E- 7,6E-0 9,56E-,00E-0,89E-05,06E-05,08E-05 00 -MN-,0757,0764,E-,60E- 7,6E-0,E-,00E-0,94E-05,8E-05,7E-05 400 -MN-,406,407,0E-,78E- 7,6E-0,76E-,00E-0,9E-05,E-05,E-05 600 -MN-,6055,6056,54E-,E- 7,6E-0,E-,00E-0,9E-05,E-05,E-05 800 -MN-,8090,8094 4,80E-,9E- 7,6E-0,87E-,00E-0,90E-05,08E-05,0E-05 000 -MN-,0049,00497,76E-,0E- 7,6E-0,8E-,00E-0,94E-05,8E-05,7E-05 00 -MN- 00 00,008 6,7E-,E- 4,00E-0 7,56E-,00E-0,45E-05,4E-05,8E-05 400 -MN- 400 400,00 7,87E-,E- 4,00E-0 8,89E-,00E-0,47E-05,46E-05,87E-05 600 -MN- 600 600,04,46E-,E- 4,00E-0 6,40E-,00E-0,4E-05,8E-05,79E-05 800 -MN- 800 800,046,E-,E- 4,00E-0 4,4E-,00E-0,40E-05,E-05,75E-05 000 -MN- 000 000,045 9,5E-,E- 4,00E-0 4,46E-,00E-0,40E-05,E-05,74E-05 00 -MN- 00 00,050,0E-,E- 4,00E-0,7E-,00E-0,6E-05,5E-05,67E-05 400 -MN- 400 400,08,8E-,E- 4,00E-0 5,07E-,00E-0,40E-05,E-05,74E-05 600 -MN- 600 600,00 4,5E-5,E- 4,00E-0,59E-0,00E-0,6E-05,68E-05,0E-05 800 -MN- 800 800,0,58E-,E- 4,00E-0 9,47E-,00E-0,49E-05,48E-05,89E-05 4000 -MN- 4000 4000,07,9E-,E- 4,00E-0,5E-0,00E-0,79E-05,94E-05,40E-05 400 6,5 MN- 400 400,0,85E- 6,00E-0 7,E-,E-0,00E-0,4E-05,E-05 4,64E-05 4400 -MN-, 6,5 MN- 4400 4400,6,4E- 6,00E-0 7,E-,07E-0,00E-0,5E-05,E-05 4,65E-05 4600 -MN-, 6,5 MN- 4600 4600,088,5E- 6,00E-0 7,E-,59E-0,00E-0,5E-05,04E-05 4,8E-05 4800 -MN-, 6,5 MN- 4800 4800,089,4E- 6,00E-0 7,E-,09E-0,00E-0,08E-05,96E-05 4,7E-05 4950 -MN-, 6,5 MN- 4950 4950,4 9,86E- 6,00E-0 7,E-,E-0,00E-0,9E-05,7E-05 4,57E-05 5000 -MN-, 6,5 MN- 5000 5000,,E- 6,00E-0 7,E-,7E-0,00E-0,7E-05,05E-05 4,40E-05
The procedure selected n ths odel s very safe wth respect to the best easureent capablty. In the case of any ndvdual uncertanty contrbutons, the axu values of an ndvdual result were used for all easureent values of a easureent. When several transducers were used, the worst result was calculated. The conservatve character of the uncertanty odel s also shown by the de En values n Table. They le very clearly below. After the copleton of last coparsons, the 5 MN force standard achne started calbraton servce n Noveber 008 as a natonal standard up to 5 MN wth an uncertanty below. 0-4. The sall uncertanty of the achne's force realzaton was pressvely shown by coparson easureents wth the MN and 6,5 MN force standard achne, usng hgh-precson transfer standards. The results are shown n Fg.. ast but not least, the revsed faclty offers, n the range up to 5 MN, twce as any force steps as the 6.5 MN faclty. ong te easureents also approve the stablty of the achne. The dfferences of repeated coparsons to the MN force standard achne are uch lower than the typcal devatons between these two achnes. The short te stablty durng a sngle easureent s assured by a regulated ol-coolng syste wthn the pressure-ppe fro the hydraulc aggregate to the achnes cylnders, whch keeps a stable teperature of C. Table. E n values for an assued overall uncertanty of. 0-4 for a k-factor of. x x Devaton W E n 0,9996 0,9996-0,000% 0,00% -0,0 0,799888 0,799898 0,00% 0,00% 0,,9985,9984 0,0000% 0,00% 0,00,59970,59979 0,0005% 0,00% 0,05,999598,999605 0,0004% 0,00% 0,04,0757,0764 0,0005% 0,00% 0,05,406,407 0,0007% 0,00% 0,07,6055,6056 0,0006% 0,00% 0,06,8090,8094-0,000% 0,00% -0,0 000 000,07 0,004% 0,00% 0, 00 00,008 0,0004% 0,00% 0,04 400 400,00 0,0004% 0,00% 0,04 600 600,04 0,0009% 0,00% 0,09 800 800,046 0,007% 0,00% 0,6 000 000,045 0,005% 0,00% 0,5 00 00,050 0,006% 0,00% 0,5 400 400,08 0,0008% 0,00% 0,08 600 600,00 0,0000% 0,00% 0,00 800 800,0 0,0006% 0,00% 0,06 4000 4000,07 0,0007% 0,00% 0,07 400 400,0 0,006% 0,00% 0,6 4400 4400,6 0,009% 0,00% 0,8 4600 4600,088 0,009% 0,00% 0,9 4800 4800,089 0,008% 0,00% 0,8 5000 5000, 0,00% 0,00% 0, 4. OPERATING PRINCIPE OF THE AUTOMATIC CONTRO In the past, the achne was controlled anually, whch eans that the ol nflow that s requred for establshng a equlbru of the pressure balance between the sde of the ass and the sde of the force transducer was set by hand. Ths procedure was extreely dffcult due to the fact that the control behavour of the achne s hghly nstable. The hydraulc aggregate s ated below the achne. Due to the heght of the constructon, long hydraulc ppes are necessary. At the sae te, the two pston/cylndersystes exhbt a relatvely large gap and the achne s very flexble thanks to the fact that s was desgned as a long, sl constructon (the a beng to acheve only sall copulsorly appled transverse forces. Although these factors bear any advantages as regards the achevng of a sall total uncertanty of the achne, they have the dsadvantage of akng a fast and stable autoatc control dffcult. For the control prncple, dfferent ethods were studed but wth none of the was t possble to acheve - as a sngle prncple - a stable and suffcently exact regulaton. A very satsfactory soluton was then acheved by a cascade whch conssted of (a an ol pressure control, (b a resdual force regulaton on the easureent sde and (c a poston control whch was desgned n a very coplex way. If a load step s to be trggered, frst of all the new ass constellaton for the force step s coposed. If there s already a force on the force transducer, an ol pressure control antans the force on the workng cylnder wth an accuracy of 0.0%. Wth the sae ol pressure fro the last load step and the new ass stack cobnaton for the next load step, the easurng cylnder decreases or ncreases towards two ltng bearngs between whch the easurng cylnder can ove by. Subsequently, the ol pressure s altered n the drecton of the next load step untl the dfference to the nonal force aounts to only 0 kn. In ths state, the selected asses and the easurng cylnder s - va a couplng eleent - brought nto a poston centrally between the two ltng bearngs. The force whch s needed for ths s easured by eans of a sall force transducer. In order to avod nfluences fro transverse forces whch are caused by unavodable anufacturng and algnent tolerances of the easurng cylnder (whch s approxately long and rotates durng the easureent wth one revoluton every 0 seconds and fro the ass stack appled to t, a transverse-force-copensated force transducer s used, as well as a specal flexure pvot whch s not able to transt any relevant torques to the transducer. After the resdual force transducer, ated at the easureent sde, has been coupled, ts sgnal s used as control quantty n order to reduce sae to zero. If ths s the case, the ol pressure generates exactly the buoyancy force for the load step selected on the ass stack,.e., after the hydraulc aplfcaton, the exact calbraton force on the force transducer. Orgnally, ths control process was supposed to be suffcent to acheve a sall uncertanty of. 0-4. Due to varous uncertanty nfluences - anly n the easureent of the resdual force - t was not possble to reduce the reproducblty below a three tes larger value. Therefore, a thrd control step was ntegrated nto the achne. After settng the ol pressure va the resdual force
copensaton as descrbed before, the force transducer used for ths s now beng decoupled agan. The asses and the easurng cylnder, whch s now freely pendng agan, would ove - accordng to the sall devatons resultng fro the uncertantes of the process descrbed before - aganst the upper or the lower ltng bearng. Now, a path control s trggered. The poston of the easurng cylnder s easured at two ponts - whch, n order to avod any rotatory effects, le opposte of each other. A coplex control algorth ensures that after release, the vety of the easurng cylnder s set to zero. Thereby, the stable regulaton of the achne turned out to be a dffcult task. Especally for the deceleraton (whch should be as fast as possble of the ass-stack syste and the easurng cylnder after ts release, fast control paraeters of the PID syste were necessary - whch would norally lead to a resonance. To avod ths, the control paraeters change contnuously, strongly decreasng n the frst seconds. Ths allows a fast, safe and very precse settng of the requred hydraulc pressure. In the range fro 00 to 5000 kn, the control varatons at the output quantty are saller than 0.00%. Attepts to operate the achne wthout the resdual force copensaton connected upstrea, and only by eans of the poston control, turned out to be nacceptable. Strong, fast control paraeters lead to a resonance and to a behavour whch s, as a atter of prncple, unstable. Weak and slow control paraeters are - n contrast to ths - stable, but they are - by far - too slow. The cascade coposed of ol pressure, resdual force and poston control s thus the ost precse and the absolutely safest ethod. At the sae te, ths ethod s relatvely fast: wthn only 50 seconds, a new load value s beng trggered. The te sequence can, however, also be altered n the control progra. For exaple, durng the ass lnkup, the achne was adapted exactly to the te response (65 seconds durng the load step change of the MN n order to nze any possble nfluences of the creep behavour of the transfer transducers. Devaton n 0-5 0 8 6 4 0 - -4-6 -8-0 C 5MN n6,5 MN buld up syste T MN ASMW and C 5MN ASMW C C MN n -MN- MN C MN n -MN- T MN n -MN-KNME C MN n -MN- C 5MN n 6,5 MN 0 000 000 000 4000 5000 Force n kn Fg.. Results of a coparson between the 5 MN and the MN and 6,5 MN. 4. CONCUSIONS PTB s new 5 MN enables low uncertantes n a strongly deanded range of force calbraton. The achne works autoatcally and enables any possbltes for further nvestgatons. It could be possble to reduce the uncertantes to a lower value than the now naed 0-4. Caused by the flexble outlne of the constructon, the achne shows extreely low rotatonal devatons durng a calbraton. The autoatc control stll has potental to enable even lower uncertantes. In addton wth the lnkup of the ass stack syste wth also lower uncertantes than average for a achne of that type, further nvestgatons wth addtonal buld up systes and nternatonal coparsons wll show the perspectve for an even ore precse, saller uncertanty budget of the achne. ACKNOWEDGMENTS The authors acknowledge the copanes of WPM n epzg and Dol n Munch for ther part of the successful odernzaton of the 5 MN. REFERENCES [] Sawla, A. Uncertanty scope of the force calbraton achnes Proc. of the IMEKO XVI World Congress, Venna, Austra, 000, Volue III, TC-, pp. 5-57.