附錄 外文資料 TEMPERATURE CONTROL P. H. J. Ingham Marketing Manager ,Eurotherm Ltd,Worthing,Sussex,UK SUMMARY Commercial plastic materials are organically based and are therefore heatsensitive .Accurate temperature control of melt processes such as injection moulding is therefore necessary if problems caused by thermal degradation are to be avoided. The injection moulding process is considered form a temperature controlriewpoint and some of the control methods or techniques are described.since it should not be forgotten that good temperature control can lend to materials and energy savings. 1 INTRODUTION The injection moulding process is concerned with the efficient conversion of plastics raw material into moulded product of acceptable standards.Some of ths parameters which determine acceptability are weight,dimensions,colour and stenght,all of which can be affected by the conditions under which the material is processed.Having established by the conditions for thwese parameters so as to deermine acceptability,limits can be set for the conditions under which the material is processed.One of the most important parameters contributing to the correct operation of an injection moulding machine is temperature.All plastics materials can be correctly processed only within a certain range of temperatures which varies from materialFor some mateials and mould types the band isvery small and for others it can be quite wide. Any attempt to define the limits within which the product is acceptable determines the need for some form of control.There are a number of types of control which,if applied correctly,can lead to adequate performance.Significant material and energy savings can be achieved by correctly pplying the right type of control equipment.The reliability of the system and the degree of operator supervision required also depend very largely on the balance struck between initial cost and performance. It is the purpose of this chapter to examine the injection moulding machine from a temperature control viewpoint and to outline some of the control methods can be used ,together with advantages and disadvantages. 2 THE PROCESS 2.1 Machine Zoning From a control viewpoint,an injection moulding machine consists of a number of zones (each equipped with a means of measauring the temperature) and a controller,which compares the measured value of the set-point and controls the heat input to the zone in such a way as to remove any different between the heat input to the zone in such a way as to remove any difference between the tow. Yu dividing the machine into a number of zones the different temperature requirements of different zones and their different heat input needs can most easily be met (Fig.1). For this purpose a typical small machine may have three or four barrel zones and a nozzle one. The zones nearest to the material feed hopper are where the plastic is melted and thus require fairly large heat inputs. However, in the zones hearest to the nozzle, the heat produced, by the rise in pressure needed to force the plastic into the mould, means that relatively little additional heat input is requied when the machine is running. Indeed, if the machine cycle very short, with some materials it may be that more heat is generated than required to maintain the temperature, which will then rise uncontrollably mless some form of additional cooling is applied. 2.2 Thermocpuple Location Considering again the barrel zones:these consist of a metal arrel with wall thickness sufficient to withstand the high pressures produced during the mjection cycle. The most common form of heating is electrical and is ipplied using band heaters strapped around the barrel (Fig.2). A controller of any kind can only control the temperature at the point of measurement. Ideally this will be as deep into the barrel wall as possible, since it is the temperature of the plastic which is required and not that of the barrel. Plastic is a poor thermal conductor and depending on whether the net heat dow is into or out of the plastic, a thermocouple deep into the barrel wall will register a temperature above or below the actual temperature. If the measuring element is shallow or on the barrel surface, the difference between the measured and actual melt temperatures can be very large. For any given conditions of operation there will be a more or less fixed difference between the melt and measured temperatures and acceptable produce may be produced. If ,however, the conditions, e.g. machine speed or ambient temperature, change, this may give rise to a melt temperature which does not result in the production of acceptable product. It is therefore important to place the thermocouple as close to the melt as possible , i.e. deep the barrel. 2.3 Temperature Overshoot The resultant system of an electrical band heater strapped around a thick walled barrel with a deep thermocouple is typical of most plastics processing machinery and present a number of control problems. Not only must stable control be achieved during normal running of the machine but acceptable start-up performance must also be achieved. The machine must be brought to its normal operating temperature as quickly as possible and preferably with no overshoot. (Overshoot is said to occur if the temperature is rising or falling at such a rate as it reaches set-point that it does not stop there but continues past by some amount before returning towards set-point again; see Fig.4.) The basic cause of temperature overshoot in the system is multiple heattransfer lags, i.e. where the heat generated electrically first raises the temperature of the heater thermal mass and is then conducted from the second thermal mass to a third and so on, until the heat reaches the point of measurement which, as stated already, is as near as possible to the point in the process to be controlled. In the simplest cast of multiple heat transfer only two thermal masses would be significantly involved, namely those of the heater and the load. If the thermal mass of each is about the same, this tends to represent about the worst case for overshoots (and hence controllability). Poor heat transfer from heater to load worsens the situation, since the heater temperature (during start-up, for example)can then become very much higher than the load temperature; when the power to the heater is cut off the final temperature reached (ignoring heat losses and assuming equal thermal masses for heater and load) will be the mean of their respective temperatures at the instant when the power is cut off. Thus ,the overshoot in load temperature increases as the heat transfer becomes worse. A particularly bad case of overshoot (and controllability) occurs where heat is transferred through a considerable thickness of heat-conducting material. This is exactly the situation which is presented by an injection machine barrel with deep set thermocouple. This sort of heat transfer represents in effect an infinite order multiple heat transfer: several minutes can elapse between switch-on of power and a significant change in thermocouple temperature. In fact the response has almost the appearance of a delay (i.e. transport lag ) although there is really a considerable difference between this heart-transfer lag and a true delay. During the time of the heart-transfer lag, heat is being fed into the barrel, so that even if the source of heat were switched off at the instant the deep thermocouple began to respond, the thermocouple temperature would continue to rise as the heat energy already fed in distributed itself evenly throughout the thickness of the barrel wall. A large part of the total lag can in practice be caused by the heart-transfer lag which occurs with a resistance heater. From the heater element thermal mass, via electrical insulation, to the outer surface of the barrel. For the lag through the barrel wall(or for any similar from the heat transfer) doubling the heart-transfer distance results in four times the lag. Iron, from which most injection machines are made, is a rather poor material for heat transfer: for example similar lag are obtained in aluminium and iron when the distance in aluminium is five times greater. 3. METHODS OF CONTROLLING TEMPERATURE 3.1 Measuring the Temperature The first item in the control system to consider is the measuring element, of which there are tow basic electrical types: active and passive. The active type are thermocouples. There are formed by the junction of tow dissimilar metals and give an output voltage proportional to the difference in temperature between the thermocouple and the point of measurement (Fig.3). The fact that the millivolt output of the thermocouple in relation to temperatures is non-linear and that it depends on a stable reference temperature for comparison purposes are factors ， Which must be taken into account in the controller. Thermocouples are very robust mechanically. (This is an obvious advantage in the environment of the moulding shop.) They also exhibit good repeatability from example to example of the same type. The two most common types used in plastic processing are both base metal thermocouples and these are nickel chrome/nickel aluminium (Type K) and iron/jconstantan (Type J). The passive types rely on having a resistance which varies with temperature in a known manner and thus, when fed from a constant current upon temperature. Such elements do not require a reference temperature to be generated by the controller. The commonest are the platinum resistance thermometer (which occupies a larer volume than a thermocouple and is more fragile)and the thermistor(which operates on the same principle and has the same disadvantages). The thermocouple is by far the most common measuring elcment used in practice. The siting of the thermocouple will depend upon the degree of control required, as will the choice of controller. 3.2 ON/OFF Control The simplest form of controller provides ON/OFF control of load power. The measured temperature is compared with the set-point and if it is too low, power is applied to the load; if it is too high the power is switched off. In practice there will be a small amount of hysteresis in the controller (mainly so that spurious noise signals on the thermocouple and effects due to mains regulation should not result in rapid ON/OFF chattering of the load power control relay). If the thermocouple and heater are in very close proximity, i.e. there is no appreciable lag, the temperature will cycle with an amplitude somewhat in excess of the controller hysteresis and with the natural period of the system. There will inevitably be some overshoot on start-up because full power will be applied to the load until the set and actual temperatures become equal and any stored energy in the heater will continue to be transferred to the load even after switch-off. It can be seen that if the thermocouple is deep in the barrel (thus measuring the melt temperature more closely) the system lags will be considerably increased and the temperature cycling will be of a longer period and will become much larger. Similar comments apply to the start-up overshoot. Thus ,in the least demanding circumstances, an ON/OFF controller with a shallow thermocouple may give acceptable results. However, with the large heaters required to give short start-up overshoot will probably be unacceptable for all but the least demanding situations and will be worse if account is taken of correct siting of the thermocouple. The natural period of the system results from a combination of heater power and location, sensor location, and the thermal mass of the system. 3.3 Proportional Control (P only) If we take an ON/OFF controller and force the switching of the output within the controller itself (with variable mark: space ratio)at a rate which is higher than the natural period, then we have proportional control. As the measured temperature approaches the set temperature, the relay will switch off(for a short time) the power supplied to the load. This point, at which just less than full power is applied to the load, is the lower edge of the ‘proportional band’. As the actual temperature approaches the set temperature more closely, less and les power is applied to the load until, when the two become equal, the power input is zero. It is general for the proportional band to be downscale of the set-point, i.e. at set-point the power fed to the load is zer.. The proportional band is usually defined as a percentage of the controller set-point scale span. Since the power applied to the load is proportional to the error or difference between actual and measured temperature (a so-called error-actuated system),it follows that if any power is required to maintain the temperature there must be some error in the system. This error is known as offset or droop (Fig.5). Since, on start-up, the load power will first be switched off at a temperature below the set-point, the resultant overshoot will be reduced. With a sufficiently large proportional band and sufficiently rapid cycling of the output power (compared to the system’s natural frequency) the oscillations in temperature will cease eventually. However, this does not necessarily mean that there will be no sart-up overshoot in temperature, but only that the subsequent oscillation will decay to zero amplitude. 英文翻譯 注塑模的溫度調(diào)節(jié)系統(tǒng) 商用塑料是最常用的，但它是熱敏感性材料。如果說因熱引起的問題是可以避免的，那么象注塑模中熔化過程中精確的溫度控制就是有必要的。] 從溫度控制的觀點和一些控制方法和技術(shù)的角度來考慮（這些方法和技術(shù)因不應(yīng)忘記而被敘述），好的溫度控制能節(jié)約和熱能。 一、介紹 注射模過程曾引起一次會議的討論，這次會議為模制產(chǎn)品的塑料原材料制定了可行性標(biāo)準(zhǔn)。一些可行性參數(shù)是重量，尺寸，顏色和強(qiáng)度。所有這些參數(shù)都受材料制造環(huán)境的影響。為了決定其可行性，為這些參數(shù)已經(jīng)建立了相應(yīng)的公差。對注射機(jī)的正確操作起作用的眾多參數(shù)中，最重要的一個參數(shù)是溫度，所有的塑料產(chǎn)品的制造都只有在特定的溫度范圍內(nèi)。這個特定的溫度范圍因材料而異。一些材料的這個溫度范圍相當(dāng)寬，而另一些材料的這個范圍卻相當(dāng)窄。 為使產(chǎn)品在允許溫度限制范圍內(nèi)，需要某些形式的溫度控制。如果應(yīng)用正確，這里有大量的類型能導(dǎo)致正確控制形式的操作。通過正確的應(yīng)用控制設(shè)備。能節(jié)省貴重的塑料和能量。系統(tǒng)的現(xiàn)實性和操作者監(jiān)管要求的程度，也很大程度上依賴于最新消耗，運輸消耗，工作費用三者之間的平衡。 這章的目的是從溫度控制的角度來檢查注射模具和列舉一些常用的溫控方法以及其優(yōu)點。 二、 過程 2·1 模具的分類 從控制的角度來說，一個注射模具由許多分區(qū)和一個控制部分組成（每一個分區(qū)有一種測量溫度的方法），控制器比較兩者之間的不同測量價值和控制兩者之間的不同，而用某種方法輸入到這個分區(qū)的熱移走。通過劃分模具的分區(qū)，能使這些分區(qū)更容易認(rèn)識，不同的分區(qū)，要求有不同的溫度和不同的熱輸入（如圖1）為了達(dá)到這個目的，一個典型的小模具就可以有3～4個桶型區(qū)和噴管區(qū)。這些離主流道襯套最近的區(qū)域是塑料要求熔化的地方。因此要求有相當(dāng)大的熱量進(jìn)給。然而，在離主流道襯套最遠(yuǎn)的澆口處，通過增加注射壓力，使塑料和澆口之間產(chǎn)生摩擦熱。這意味著，當(dāng)模具在工作時只需要相當(dāng)小的熱量輸入。如果機(jī)器的循環(huán)周期非常短。某些材料在制造過程中比被要求的熱量產(chǎn)生更多的熱量，為了保持溫度，就需要采用某些形式的冷卻方式應(yīng)用。 2·2 熱電偶的安裝 再考慮這些桶型區(qū)：一個型腔應(yīng)具有足夠的壁厚。用以承受足夠的壓力。最平常的加工方法是電加熱和使用一個帶狀的加熱片貼在型腔周圍（如圖2），在任何類型的一個控制器都只能控制一個點的測量溫度的測試，而且盡可能貼近型腔。因為我們需要的是塑料的溫度，而不是型腔的溫度，塑料是熱的不良導(dǎo)體。依靠純熱進(jìn)去塑料，如果熱電偶安放在型腔的表面或非常淺，那么測量值和實際值之間將會有非常大的差異。 任何給出的操作環(huán)境都或多或少的存在實際值和測量值之間的差異。然而如果環(huán)境變化，如模具的運動速度和周圍的環(huán)境溫度變化，這都可以影響到工件的熔化溫度。因此，熱電偶的安裝位置要盡可能的靠近型腔的內(nèi)壁。 2·3溫度過調(diào)量 一個具有一個熱電偶的加熱片貼在一個深孔型腔的壁上。它的合模系統(tǒng)是最典型的塑料加工機(jī)械，而且存在著大量的控制問題，不僅在正常的模具工作期間必須完成穩(wěn)定的控制，而且可行的合理的初始操作也必須完成機(jī)械可以在不用調(diào)節(jié)時盡可能完美而迅速地使它達(dá)到正常的操作溫度（如果溫度上升或下降，以某一頻率。就是說它經(jīng)過那點，但不停留在那點，而是在它返回那點時繼續(xù)通過一定數(shù)量的點。在這種情況下，過量調(diào)節(jié)就出現(xiàn)了。如圖4） 在系統(tǒng)中引起過量調(diào)節(jié)的基本原因是，多個熱傳導(dǎo)滯后等產(chǎn)生的殘余熱量。首先，引起受熱物體的溫度上升，然后，傳遞給第二個受熱物體，同時使第二個物體溫度上升，然后從第二個受熱物體傳遞給第三個受熱物體。以次類推直到熱在傳遞過程中達(dá)到控制溫度的點附近。 舉一個最簡單的多個熱傳遞的例子，如果兩個受熱體，如果每個受熱體都是一樣的，那將是過調(diào)量中最糟的。一種情況，沖加熱到裝入的差的熱傳遞使環(huán)境變糟，因為加熱溫度（如在開始時的溫度）。將使最終裝入溫度遠(yuǎn)高于其本身。當(dāng)加熱電源切斷時，最終溫度就達(dá)到了。（忽略溫度損失和假設(shè)加熱熱量和吸收熱量相等）。這將意味著最終電源切斷時，最終各方面的溫度。因此，過調(diào)量作為過調(diào)量作為熱傳遞在裝入溫度上升時變地更糟。 在特別糟的過調(diào)量（可控制）的情況出現(xiàn)在熱傳遞通過熱導(dǎo)體材料的深處，這是實際的環(huán)境。這個環(huán)境是一個具有深的安裝電熱偶的注射模具環(huán)境。這套熱傳遞系統(tǒng)抽繪一個無限次續(xù)的多熱傳遞系統(tǒng)的影響。在打開電源和在熱電偶中的一次重要轉(zhuǎn)變之間需要幾分鐘的時間。實際上，這反映的是一種延時的表現(xiàn)（如傳導(dǎo)滯后），雖然熱傳導(dǎo)滯后和真正的延時之間存在著差異，在熱傳導(dǎo)滯后和真正的延時之間存在著差異，在熱傳導(dǎo)滯后的時間中，熱進(jìn)給到型腔，以至于熱源被切斷的瞬時深的熱電偶開始反應(yīng)，當(dāng)熱能已經(jīng)進(jìn)給通過整個型腔壁后來完全地分配本身。 總的滯后的大部分，可以是由于發(fā)生在熱阻傳導(dǎo)體的熱傳導(dǎo)滯后引起，熱阻傳導(dǎo)體從熱的基本發(fā)熱體，經(jīng)過電隔離在型腔外表，因為滯后通過型腔壁（或任何一個類似的熱傳導(dǎo)）兩倍的熱傳導(dǎo)距離而產(chǎn)生了四倍的滯后。大多數(shù)注射模具制造用的鋼材對熱傳導(dǎo)是相當(dāng)差的材料。舉一個簡單的例子：當(dāng)在鋁中的距離比在鐵中大五倍時。在鐵和鋁中能得到相同的熱滯后。 三、 溫度控制的方法 3·1溫度的測量 在控制系統(tǒng)中，首先要考慮的一條是測量的元素，它有兩種基本的電子測量類型：主動的和被動的類型。 主動類的是熱電偶，它由兩種不同金屬片和一個外部電壓組成。這個外部電壓與熱電偶和測量點之間的不同溫度相稱（如圖3）；熱電偶的毫伏輸出電壓與溫度不成線性關(guān)系，它依賴一個作為比較目的的穩(wěn)定的參考溫度，這一事實都是在控制器里必須考慮的因素，熱電偶具有相當(dāng)強(qiáng)的機(jī)動性（這在模具工廠的環(huán)境中是相當(dāng)有利的）。這些因素也表現(xiàn)好的重復(fù)性。從例子到相同的類型的例子，兩個最常用在塑料加工過程的例子都是金屬熱電偶的基本組合材料，它們是鎳鉻/鎳鋁合金（類型K）和鋼/銅合金（類型J）。 無源類熱電偶，存在一種阻力，這種阻力使溫度不同于眾所周知的那種方式。因此，當(dāng)在恒流電源的作用下，這種阻力將產(chǎn)生電壓，這個電壓依賴于所通過的材料的溫度。最常用的是鉑阻熱電偶（這種熱電偶比以前講的普通熱電偶具有更大的容量，并且更容易碎。）和熱敏電阻（它是用同樣的原理進(jìn)行工作具有同樣多的不利條件）。 熱電偶是在實踐中被大量使用的最常用的測量工具。熱電偶的定線將依賴于要求控制的度數(shù)和所選的溫度控制器。 3·2控制器的開關(guān) 控制器的最簡單的形式提供負(fù)載電源開關(guān)的控制，測得的溫度與安裝點比較，假如溫度太低，負(fù)載電源將參與工作，假如溫度太高，負(fù)載電源見被切斷，在實際中，在控制器中有一些磁滯現(xiàn)象。如果熱電偶和加熱器非常接近，那么這就不存在滯后，溫度將以某種振動進(jìn)行循環(huán)。這個振幅是由控制起的滯后和系統(tǒng)的自然周期引起，因為全功率的電源在要求的溫度和實際溫度相等之前一直提供負(fù)載，所以在開始時有一定的過調(diào)量是不可避免的。很明顯，如果熱電偶在型腔壁的深層（因此測量的熔化溫度更接近）。系統(tǒng)的滯后增大，溫度的循環(huán)周期將變長，振幅將變大，也同樣在開始時有一個過調(diào)量。 因此，一個具有線的熱電偶開/關(guān)控制器可以得出所接受到的結(jié)果，這是起碼的要求。然而具有大的熱電偶的開/關(guān)控制器要求有一個更短的啟動時間。如果計算考慮了這個熱電偶的正確安放位置，那么這個啟動時間過短將可能是對于所有控制器來說是不接受和更糟的。除這起碼的要求。 這套系統(tǒng)的自然時期來源于一個熱電偶能量與位置的聯(lián)合作用，傳感器的位置和系統(tǒng)的熱量集中區(qū)域三個因素。 3·3比例的控制（僅僅是P的控制） 如果我們使用一個開/關(guān)控制器，并且迫使輸出量轉(zhuǎn)換。在控制器內(nèi)部本身有一個頻率，這個頻率高于自然時期的，然后我們將要進(jìn)行一個比例的控制問題。當(dāng)測量的溫度接近安放點的溫度時，繼電器將在短時間內(nèi)切斷提供負(fù)載電源，在比最大電源電壓少一些的這個點是比例帶的最低邊緣，當(dāng)實際溫度接近安放點的溫度時，越來越少的電源電壓進(jìn)給量，直到兩者完全相同時，電源輸入量將變成零。總的一句話來說，對于比例帶到安放點呈降低的比例趨勢。例如在安放點的電量進(jìn)給為零。 比例帶的定義就是一個控制器安放點的范圍段的一個百分率。因為電源負(fù)載的誤差是成比例的，或是實際溫度與測量溫度之間存在著差異（一個所謂的誤差一個實際系統(tǒng)），這產(chǎn)生的后果將是假如任何電源要求保持溫度，這將使在系統(tǒng)中產(chǎn)生某些錯誤，這個誤差就是眾所周知的偏差和下降（如圖5）。然而在開始上升階段，在溫度還低于安放點時，負(fù)載電源將被關(guān)掉，短期內(nèi)的結(jié)果將降低，用一個足夠大的比例帶和足夠快的外部輸出電壓的循環(huán)（與系統(tǒng)本身的自然頻率相比）溫度的波動將最終停止。然而，這并不意味著這里沒有上升的過調(diào)量，而僅僅只是意味著在此以后的波動將減小到振幅為零。