8
Furnaces and Services for Metals. SONDERDRUCK · SPECIAL PRINT Metallurgical Plant and Technology Metallurgical Plant and Technology 6 November 2007 Original veröffentlicht in MPT 6/2007 Quality aspects of hydrogen annealed steel strip Quality aspects of hydrogen annealed steel strip INTERNATIONAL INTERNATIONAL

Quality Aspects of Hydrogen Annealead Steel Strip_e

Embed Size (px)

Citation preview

Page 1: Quality Aspects of Hydrogen Annealead Steel Strip_e

F u r n a c e s a n d S e r v i c e s f o r M e t a l s .

S O N D E R D R U C K · S P E C I A L P R I N T

Metallurgical Plant and TechnologyMetallurgical Plant and Technology

6November

2007

Original veröffentlicht inMPT 6/2007

Quality aspects of hydrogen annealed steel stripQuality aspects of hydrogen annealed steel strip

INTERNATIONALINTERNATIONAL

Page 2: Quality Aspects of Hydrogen Annealead Steel Strip_e

2 MPT International 6/2007

Introduction

Sheet steels annealed in state-of-the-art Bell-Type Annealing Furnaces(BAF) fulfil highest quality require-ments. The mechanical propertieshave been optimised in respect ofdrawability, for low yield strength,high elongation and high Lankfordvalues. With the same importance theBAF has to ensure best surface qualitywithout any defects. Since soft gradestend to be very sticker sensitive theavoidance of sticker marks is a majorissue of quality aspects.

Classification of stickers

"Sticker" is an overall term for theplastic deformations or flow and kinkmarks already apparent on steel stripon the temper mill during the uncoil-ing of the windings of the coil before itreaches the rolling gap. Stickers areproduced by the pressure welding ofbare metal surfaces. It has not yet beendecisively established whether thiswelding is caused by diffusion weldingprocesses between two surfaces, sinter-ing processes or other adhesion mech-

anisms. However, a number of differ-ent types of stickers have beenobserved. In some cases, the processeswhich create these defects have beenidentified.

Ridge stickers are sharply limiteddefects over the width of the strip.These stickers are mainly caused bystrip profile anomalies (such as ridges),which can cause high radial pressureswithin the windings.

Spot stickers are limited to localizedspots on the strip. The main cause ofthese stickers is the local application ofhigh pressures (by coil tongs, forexample) and the unintentional wind-ing of foreign objects into the coil,which become apparent when the coilis unwound.

Edge stickers, as the name implies,occur at the edge of the strip. They aresharply limited flaws on the strip edgecaused by relative movement betweenthe windings of the coil and convectorplates or charge carriers as a result ofdifferences between the heating andcooling rates of the coil and these parts.Damage caused by coil handling, such

In the past, the bell-typeannealing process has often

been seen as the sole cause ofstickers although there are anumber of different types of

stickers caused by many differ-ent factors. This article discussesthe various types of stickers, themechanisms that produce them

and the precautions which can be taken to prevent theiroccurrence. Throughout theworld, there are companieswhich suffer losses of up to10% as a result of stickers.Leading rolling mills which

have focussed on the problemof stickers and the processes

that produce them have beenable to reduce the average

share of stickers to less than 1%.Stickers have therefore lost the

prominence among surfacedefects that they had in

the 1990s.

Quality aspects of hydrogen annealed steel strip

Dr.-Ing. Peter Wendt, Director of Sales;

Dipl.-Ing. Frank Maschler, Senior Process

Engineer; Dr.-Ing. Peizhong Wang, Senior

Sales Manager, LOI Thermprocess GmbH –

LOI Italimpianti, Essen, Germany

Contact: www.loi-italimpianti.com

E-mail: [email protected]

Figure 1. Modern batch annealing plant supplied by LOI Thermprocess GmbH

Reprint from “Metallurgical Plant and Technology“ 6 (2007), pages 54–162© 2007, Verlag Stahleisen GmbH, Düsseldorf

Page 3: Quality Aspects of Hydrogen Annealead Steel Strip_e

as positioning on a surface which is notlevel, may also cause edge stickers.

Edge stickers can be reduced byusing convector plates with a highermanganese content, which are harder.The design of the convector plates [1]and charge carriers is also important.The use of intermediate convectorplates with a carrying area of 50% ormore and machined intermediate con-vector plates can help to minimizeedge stickers on thin strip (thicknessless than 0.7 mm).

General stickers are irregular kinkedlines which take on a sickle-shapedappearance with increasing width (50to 400 mm wide) and are mainly dis-tributed over a large area in the centreof the strip, with reference to its width.They are also referred to as flowingmaterial shapes or Lüders lines (figure2).

Because of the crown of the stripand the resulting higher pressure onthe strip centre, the flaw is positionedin a large area at the centre of the stripwidth in the case of strips with an opti-mum symmetrical profile. If the crownof the hot strip, which determines theprofile of the cold strip, is too high ortoo sharp, the result may be a narrowbut very pronounced track of stickersalong the centre of the strip width.

Strips with profiles which are toolow or too flat may have very widesticker areas or sticker tracks in theouter thirds of the strip width. In thecase of strips which have a slightlywedge-shaped, asymmetrical cross sec-tion, the stickers will be located awayfrom the central position at the high-est part. The creation of general stick-ers is affected by a wide variety ofprocess parameters, some of whichhave contrary effects. The causes ofstickers and possible countermeasuresare considered below.

General stickers and theircauses

Stickers are created on the tempermill when coils annealed in a bell-typeplant are uncoiled and the forcerequired to open or separate the coilwindings is in excess of the yieldstrength of the annealed material. Inother words, general stickers dependboth on the stress applied to the coilduring unwinding and on the yield

strength of the annealed material. As aresult, "welds", between the windingsof the strip which can be separated byapplying force below the materialyield strength do not result in stickers.

Investigations have shown that theformation of sticker marks is largelydetermined by the actual value of theyield strength and by the behaviour ofthe material at stress approaching theyield strength (figure 3).

An evaluation of experience ob-tained with stickers at a German coldrolling mill over a period of severalyears shows that:- sticker frequencies are higher with

soft grades (yield strength < 200N/mm2),

- no or virtually no stickers occur withhigher-strength grades (yieldstrength > 220 N/mm2),

- no or virtually no stickers occur withsoft IF steels because of the lack of aprecisely defined yield strength.The effective local force acting on

the strip during uncoiling in thehatched area of the above diagrammust therefore be 120 - 220 N/mm2

(figure 3). This force results from theseparation force and the additionaleffects of uncoiling speed and geome-try. Higher separation forces are causedby strong adhesion between the stripwindings resulting in the "welding" oftightly packed strip surfaces. Theseadhesion forces are affected by condi-

Figure 2.

Sticker marks

Figure 3.

Effect of the yield

strength on the

sensitivity to

sticker formation

0

Influence of yield stress

Co

ils w

ith s

tick

ers

, %

28

24

20

16

12

8

4

0

Critical Not critical

120-149 150-179 180-209 210-239 240-269 >270

Yield stress, N/mm2

Yie

ld s

tress

, N

/mm

2

400

350

300

250

200

150

100

50

0Soft quality High-strength quality IF quality

Separationforce

Heat treatment

3MPT International 6/2007

Page 4: Quality Aspects of Hydrogen Annealead Steel Strip_e

tions during the annealing process(temperature, pressure and annealingtime).

The causes of these welding phe-nomena were comprehensively inves-tigated for the first time by Pawelski etal. [2]. To date, knowledge of the caus-es of stickers in Europe is still based onthis report, published in 1989. Theeffects of pressure on maximum sepa-ration force are considerably increasedat higher temperatures. At tempera-tures up to about 600 °C, practicallyno welding occurs, irrespective of thetension on the strip. This conclusion isconfirmed by practical experience,although temperatures up to a maxi-mum of 620°C are generally regardedas uncritical with regard to stickers. Ingeneral, the higher the annealing tem-perature, the less time is required forwelding the samples.

Factors affecting sticker formation

Although the pressure welds whichlead to stickers are always produced inthe batch annealing process, thedegree of welding and the resultingseverity of the stickers are determinedto a very large extent by upstream anddownstream production stages. Thefactors concerned can be divided intothose related to- upstream process stages,- the annealing process itself, and- downstream process stages.

As shown in table 1, the main fac-tors can be summarised in three cate-gories: before, during and after anneal-ing.

Steel grade. Soft, unalloyed steels ofgrades DC 04 (DDQ) and DC 05(EDDQ) are sensitive to sticker forma-tion [3]. IF steels without a clearlydefined yield strength after annealing,higher-strength structural steels, high-

strength and micro-alloyed steels areconsidered to be less susceptible tostickers.

Strip profile. The profile of the cold-rolled strip reflects that of the hot-rolled strip. A steel strip with a profileoptimized to avoid stickers has acrown of 40-60 µm. Significantlylower crowns (flatter profiles) lead toincreased sticker rates in the edge orcentral sections as a result of higherpressures. Asymmetric (for example,wedge-shaped) cross sections and crosssection anomalies such as ridges alsoresult in higher pressures and producemore intensive welding between thewindings of a coil during the batchannealing process.

Coiling tension. Coiling tension is akey factor in determining the stress inthe coil and therefore also the stickerrate. Minimal coiling tension (near tothe collapsing limit) helps to reduceoverall compressive stress during theannealing process. Lower limits to thecoiling force are set by the need toensure safe handling during removalfrom the coiler and transfer to the bell-type annealing furnace, the need for a stable coil and the avoidance ofscratches caused by slippage duringannealing and dressing. Leading coldrolling mills are in a position to reducethe coiling tension applied to a mini-mum (~ 22 N/mm2). The coiling ten-sion must be optimized for the indi-vidual coil dimensions, widths andthicknesses.

Strip roughness and topography.The roughness of the strip surface hasconsiderable impact in minimizingwelding and reducing the separationforce required. Pawelski et al. [2] havealready investigated these effects. Theresults of their experiments are shownin figure 3. The separation forcesdetermined for roughened strip sam-

ples in the temperature range from670 to 700°C were only about half thecorresponding figure for smooth sheetwhich had not been roughened priorto annealing.

Apart from the roughness value, thetexture of the surface roughness is alsoimportant. Surface irregularities whichare more pointed and have a finerstructure are more effective in reduc-ing welding than coarser structureswith a plateau configuration.

For O5 surface qualities, leading coldrolling mills in Germany aim for a pre-roughened structure with a roughnessof Ra > 1.2 µm and a minimum num-ber of peaks of Pc > 50 /cm (> 127 ppi).In order to avoid stickers, the valuesshould not fall below Ra >1.0 µm andPc > 40 /cm (> 100 ppi).

Strip cleanliness. As a general princi-ple, the cleaner the strip, the greater isthe risk of welding. The strip clean-liness of a rolled strip is normallydefined in terms of oil residues andiron fines in mg/m2. The introductionof HPH® (High Performance Hydro-gen) annealing technology significant-ly improved strip cleanliness andtherefore led to an increased tendencytowards sticker formation with other-wise unchanged conditions. High stripcleanliness, which is desirable fordownstream process stages, is there-fore counterproductive in terms ofsticker formation.

If electrolytic strip cleaning systems(ECL) are used upstream from theannealing furnace, oil deposits can bereduced to 5 mg/m2 at the same timeas forming a silicate layer of the orderof 4 - 8 mg/m2 SiO2, which forms abarrier and prevents welding betweenthe windings. An Asian cold rollingmill has reported that it was able tohalve the sticker rate (from > 6% to <2.9%) simply by optimizing the thick-ness of the silicate layer. A number ofattempts have been made and are stillbeing made to apply separating layersto the strip surface to prevent pressurewelding without any detrimentalimpact on the production process.

Tests with the addition of a silicateto the rolling emulsion for the laststand of a five-stand tandem line wereat least partially successful. If thisapproach is adopted, it is necessary toensure that the pH and stability of theemulsion are not impaired. Theremust be no adverse effect on corrosion

Heat treatment

4 MPT International 6/2007

Factors affecting sticker formation

Before annealing During annealing After annealing

Steel grade Cooling rate Uncoiling speedHot strip profile Coiling tension Heating-up Uncoiling tensionStrip roughness gradient Strip cleanliness Uncoiling geometryStrip dimensions Coil positionCoil dimensions

Table 1. Important factors affecting sensitivity to sticker formation

Page 5: Quality Aspects of Hydrogen Annealead Steel Strip_e

and welding behaviour or on down-stream production stages such as phos-phating, painting, hot-dip galvanizingor electrolytic galvanizing. A Germancustomer has reported that the stickerrate could be reduced from about 2%to about 0.2% by applying a siliconlayer of 2 - 3 mg/m2 SiO2 in the lastrolling stand [4].

The DEFOX process in which anoxide layer only effective in the criticalfirst few hours of cooling is appliedduring the annealing process, did notprove to be effective in practice andhas now been discontinued by many

rolling mills. The main reasons wereadhesion problems caused by incom-plete reduction of trivalent iron oxidecomponents in the layer applied, inaddition, this process had less effect inreducing stickers than the furnacecooling stage which it replaced.

Other possibilities which have beenconsidered or tested include the use ofQUERL separating layers containingnitrates, iron or iron oxide fines to pro-vide spacers with a thickness in theµm range between the windings. Todate, none of these approaches hasbeen widely adopted in practice.

Strip dimensions, coil dimensions.As a general principle, thinner, widerstrips and larger coil diameters tend tobe more critical with regard to stickers.Strip thickness is the key parameter inthis connection. The main reason isthat coils of thinner strip have a largernumber of windings, with an adverseeffect on radial heat transmissionthrough the coil. During the cooling

process, temperature differences aretherefore higher, resulting in moresevere thermal stress. Experience indi-cates that coils with strip thicknessesbelow 0.7 mm are especially sensitiveto stickers.

The negative impact of strip width isslightly less severe than that of thick-ness. Current statistics show that therisk of stickers already increases at stripwidths above > 1,100 mm. As theouter diameter of the coil increases,temperature differences during theheating and cooling process also rise,leading to higher stress on the coil and

a greater risk of welding between thewindings. Sticker rates are already sig-nificantly higher for coils with diame-ters above 1,600 mm.

After annealing. Stickers are detect-ed after first uncoiling after annealing.Therefore the uncoiling conditions areevident, mainly the uncoiling geome-try and the uncoiling speed. A properuncoiling geometry ensures a welldefined, constant separation point.One measure of implementation is ananti-sticker roller [3]. Because the yieldstrength increases with the speed offorce increase, high uncoiling speedabove 500 m/min, better 800 m/min,helps to reduce stickers.

Measures during annealingto prevent stickers

Strip welding is caused during theannealing of thin sheet coils in bell-type annealing furnaces. Pressure

welding is influenced directly by theannealing and cooling process or thetime-temperature plot and indirectlyby the thermal stress in the coil result-ing from the temperature gradient.The risk of sticker formation can bereduced by the following measurestaken during the annealing process.

The heating rate shall be reducedwhen a control temperature of 600 °Chas been reached. This approachreduces stress caused by heating-upand also has a positive impact on stripcleanliness. However, it should benoted that stickers caused purely byheating-up are rare. The time-tempera-ture regime has a greater effect.

Very long soaking times shall beavoided. As there is a considerable ten-dency towards welding at high tem-peratures, even relatively low stresswithin the coil can result in sticker for-mation. The soaking time must be setso that the critical coil in the stackreaches the core temperature required.

Cooling under the heating hoodshall occur for a maximum of 6-8hours. When the heating hood isremoved, the charge is naturallycooled very rapidly. By allowing thecharge to cool under the heatinghood, the cooling shock is reducedand shifted to a lower temperaturerange, where the risk of welding andsticker formation is reduced.

The same also applies to the startingtemperature for rapid cooling. At anyrate, the bypass or spray cooling sys-tem should not be started up until thecoil core temperature is lower than600°C.

StickerMod - a model foravoiding stickers

Until recently it has not been possi-ble to predict the occurrence of stickersin advance. As a result, the efficiency ofthe production process may beimpaired, for example if the charge iscooled with the heating hood longerthan is necessary for sticker prevention.

StickerMod was developed by LOIThermprocess [5, 6] as a tool for pre-dicting stickers in individual stacksand for the analysis of individualannealing cycles. It is based on theinvestigations of Pawelski et al. [2],who considered the theoretical back-ground (i.e. calculation of thermalstresses) and performed a number of

Figure 4. Concept of StickerMod

Heat treatment

5MPT International 6/2007

Page 6: Quality Aspects of Hydrogen Annealead Steel Strip_e

experiments on the influence of pres-sure, temperature and annealing time.Thermal stresses result from tempera-ture differences in the coil. To calculatethermal stresses, information on localtemperature distribution during thewhole annealing cycle is necessary.LOI's thermodynamic mathematicalmodel HeatMod calculates tempera-ture distribution during the annealingcycle with high precision and wastherefore used as the basis for theStickerMod model (figure 4). However,sticker calculations cannot be madesolely on the basis of stress values. Theeffects of temperature and duration arerepresented by the equivalent separa-tion force (ESF), which is an indicatorof the uncoiling forces required andtherefore also of sticker creation. TheESF is calculated on the basis of the

investigations by Pawelski et al. [2],who measured the separation forces asa function of stress, temperature andannealing duration. Stickers occur ifthe ESF reaches a critical value.

Parameters influencing stickers. Asmentioned above, stickers are influ-enced by a number of parameters.StickerMod can take some of theseinto account:- coiling tension,- outer diameter, width of strip,- heating-up gradient,- cold spot temperature,- soaking temperature,- cooling with heating hood,- start temperature for rapid cooling,- material properties (yield strength).

The influence of some of theseparameters is indicated below takingthe example of a 4-coil stack with thegeometry and annealing cycle data asfollows:

- coil width: 1,250 mm- outer diameter: 1,855 mm- inner diameter: 610 mm- strip thickness: 1 mm- stack weight: 94.6 t- heating-up gradient from 350°C to

710°C in 17 h- soaking temperature: 710°C- end of heating: 10 h equivalent time - cooling with heating hood: 3 h- start of rapid cooling: 450°C

Evaluation of process parameters

The results of the base geometrycalculation are shown in figure 5. As aresult of the geometry of the furnace,which will not be discussed here, thehighest gradients and therefore also

the highest equivalent separationforces (ESF) under these specific con-ditions can be seen on the secondcoil. For this reason, only curves forthe second coil are shown in the dia-grams.

The bold blue line represents thelocal ESF for coil 2 two-thirds of theway from the inner diameter to theouter diameter at the centreline (2/3position). During heating and soak-ing tensile stress (total stress > 0) at 2/3 position is in evidence. There-fore there is no increase in the ESF.Approximately one hour after thestart of cooling with the heatinghood, the core temperature becomeshigher than the edge temperature.The layers are pressed together (totalstress < 0) and pressure welding starts.Five hours after the end of soakingthe ESF reaches its maximum. Afterthis time less pressure will be expec-ted.

Outer diameter and width of coils.The higher width results in higher sep-aration forces increasing with increas-ing diameter. The reasons for thisbehaviour are the temperature differ-ences, which increase with increasingwidth and outer diameter. The basestack is marked by circles.

Cooling with the heating hood. Aheating hood is often used to preventstickers. Increasing the duration ofcooling with the heating hood reducesthe ESF by up to 18%. An increase inthe duration of cooling with the heat-ing hood beyond six hours has no fur-ther effect on the separation force, butonly increases the cycle time. As a second result, the separation forceincreases if the charge is only cooledunder the heating hood for one hour.The reason for this effect is the risingcore temperature at the start of coolingwith the heating hood. If the heatinghood is removed at maximum coretemperature, temperature differencesare maximized, resulting in maximumpressure at a high temperature level.

Rapid cooling start temperature. Thecontrol temperature at the start of rapidcooling has only a slight influence onthe separation force. A reduction in thestart temperature below 450°C has noeffect on the base stack. Above 450°C,the separation force increases by about4%. This can be explained by the geom-etry of the furnace. Due to the arrange-ment of the burners the second coil isheated up fastest. When rapid coolingis started, cooled purge gas is heated upby the first coil before reaching the second one.

Core temperature. It was expectedthat separation force would increasewith core temperature. However, cool-ing with the heating hood becomesmore effective with increasing coretemperatures. This can be explainedby the increase in the ESF not onlywith the pressure level but also withthe temperature level.

Soaking temperature. The separa-tion force decreases very strongly witha decreasing soaking temperature (upto 37%, compared to the base stack),while the total cycle time increasesonly by 7%. To make results compara-ble, all calculations were performedwith an equivalent time of 10 h.

Heat treatment

6 MPT International 6/2007

800

700

600

500

400

300

200

100

0

-100Tem

pera

ture

, °C

- St

ress

, N

/mm

2 8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

-1,000

ESF,

N

0 10 20 30 40 50 60

Time, h

Control temperatureCore coil 2Edge coil 2Temp. differenceTotal stress coil 2 2/3 pos.Sep. force coil 2 2/3 pos.

Width 1,250 mmOuter diam. 1,855 mmThickness 1.0 mm

4 coil stack 94.6 t

Figure 5.

Annealing cycle

for base stack by

StickerMod

Page 7: Quality Aspects of Hydrogen Annealead Steel Strip_e

Heat treatment

7MPT International 6/2007

StickerMod validation. To verify theStickerMod programme, calculationswere performed for 59 stacks whichhad already been annealed at a Ger-man plant. In 30 stacks one coil ofeach stack showed sticker marks, while29 stacks were annealed without anymarks. The maximum equivalent sep-aration force occurring during theannealing cycle was calculated foreach coil of all the stacks. Assuming acritical separation force of 6,000 N,80% of the coils with sticker markswere detected correctly. The calcula-tion of stacks without any stickermarks resulted in correct results for86%. For 14% of stacks without stickermarks the model calculated a stickerwarning. In view of the fact that thecalculations are based on experimentaldata with deviations of up to 1,700 Nand that these curves were approxi-mated by mathematical functions,80% represents good agreement. Fur-thermore it must be considered thatthe input variables of StickerMod are

only a part of the influence factors list-ed in table 1.

Conclusion

Although sticker occurrence isalways related to bell-type annealing,the main influences come from theupstream and downstream processsteps as well as from the annealingprocess itself. Many of these factors aresubjects of improvements, but some ofthem remain as input parameters to beconsidered in the design of the bell-type annealing process. High heatingand cooling speed is favourable for thethroughput, but they could be detri-mental in respect of stickers. Produc-tivity as a combination of quality andthroughput depends on how close andhow safe certain limits can be usedout. StickerMod is a kind of processintelligence which supports the designof annealing cycles. The resulting fur-nace cycles enable to achieve high

throughput by keeping the limitsgiven by the upstream process steps. �

References[1] Dr. Scheuermann, W.; Maschler, F.; Dr.

Wendt, P.: Aspects of HPH®-HydrogenBatch Annealing, Paper published on LOIInternational Customer Convention onHeat Treatment of Steel Strip and Wire,2004

[2] Pawelski, O.; Rasp, W., Martin, G.: Entste-hung von Bandklebern bei haubenge-glühtem Kaltband (Origins of sticker markson batch annealed cold rolled sheet), Stahlu. Eisen 109 (1989), No. 4, p. 178-184

[3] Emde, E.; Dzafic, A.; Frech, W.: Sticker-pre-vention measures in HPH BAF and inupstream and downstream process steps,Paper published on LOI International Cus-tomer Convention on Heat Treatment ofSteel Strip and Wire, 2004

[4] Henkel, Product Information, P3-percy [5] Dr. Leifgen, U.; Sinter, A.: Sticker Preven-

tion in Batch Annealing - StickerMod:Basics and Results, Paper published on LOIInternational Customer Convention onHeat Treatment of Steel Strip and Wire,2004

[6] Dr. Wendt, P.; Frech, W.; Dr. Leifgen, U.:Cold rolling defect "Stickers" and counter-measures, Heat Processing (5), Issue 2,2007, p. 127-135

Page 8: Quality Aspects of Hydrogen Annealead Steel Strip_e

Hot Formed Materials AluminiumAutomotive Parts,Tubes & Rods andother Semifinished

Products

Steel Strip

Genova, Italy Essen, Germany

LOI Italimpianti –The Leader of Furnaces for Metals

LOI Thermprocess GmbHAm Lichtbogen 2945141 Essen, GERMANYPhone +49 201 1891 1Fax +49 201 1891 321e-mail: [email protected]

LOI ItalimpiantiTorre ShippingVia De Marini, 5316149 Genova, ITALYPhone +39 010 6054433Fax +39 010 6054741e-mail: [email protected] www.loi-italimpianti.com

LOI Inc. Pittsburgh, USA

Techint Italimpianti DeutschlandDüsseldorf, Germany

LOI PolandTarnowskie Góry, Poland

LOI LiègeBelgium

LOI MoscowRussia

TEMOS MoscowRussia

LOI TianjinChina

LOI ShanghaiChina

LOI BeijingChina

HYPERTHERMMumbai, India

LOI-Wesman ThermprocessCalcutta, India

ww

w.to

-bic

om.d

e 0

9611

-08