Abstract

Reliable cleaning processes are an absolute prerequisite to ensure the production of safe foods. In particular, the cleaning of heat exchangers is a crucial process step since temperature induces the formation of strongly bound food-borne deposits. Aiming on relationships between soil-specific physicochemical properties and cleaning behavior, four native starches of different botanical origin (maize, waxy maize, potato, wheat) were selected for plane channel flow (similar to the flow regime in plate heat exchangers) and falling film cleaning experiments. In addition, their chemical composition and swelling behavior was analyzed. Different cleaning mechanisms could be observed, whereby diffusive dissolution was identified as dominant cleaning mechanism for most starch deposits in both cleaning experiment setups. The cleaning of such soils is characterized by an initial reptation time, where the cleaning fluid penetrates the soil and no cleaning occurs, followed by a period with constant cleaning rate. Since these cleaning parameters where overlapped by swelling-induced processes, a correction model was used to identify the unaffected cleaning progress. The cleaning and swelling-induced parameters were influenced by the origin of the starch as well as by temperature and sodium hydroxide concentration of the cleaning fluid. Multivariate statistics was helpful to unravel interrelationships and interdependencies between results of cleaning experiments, and compositional and physicochemical properties.

摘要

可靠的清洁过程是确保生产安全食品的绝对先决条件。特别是，热交换器的清洁是一个关键的工艺步骤，因为温度会导致形成强结合的食源性沉积物。针对土壤特有的理化性质与清洁行为之间的关系，选择了四种不同植物来源的天然淀粉（玉米、糯玉米、马铃薯、小麦）用于平面通道流（类似于板式换热器中的流态）和降膜清洁实验。此外，还分析了它们的化学成分和溶胀行为。可以观察到不同的清洁机制，因此在两种清洁实验设置中，扩散溶解被确定为大多数淀粉沉积物的主要清洁机制。此类污垢的清洁以初始蠕动时间为特征，此时清洁液渗入污垢且不会发生清洁，然后是一段具有恒定清洁速率的时期。由于这些清洁参数与溶胀诱导过程重叠，因此使用校正模型来识别未受影响的清洁过程。清洁和溶胀诱导参数受淀粉来源以及温度和清洁液氢氧化钠浓度的影响。多变量统计有助于解开清洁实验结果与成分和物理化学性质之间的相互关系和相互依存关系。由于这些清洁参数与溶胀诱导过程重叠，因此使用校正模型来识别未受影响的清洁过程。清洁和溶胀诱导参数受淀粉来源以及温度和清洁液氢氧化钠浓度的影响。多变量统计有助于阐明清洁实验结果与成分和物理化学性质之间的相互关系和相互依赖性。由于这些清洁参数与溶胀诱导过程重叠，因此使用校正模型来识别未受影响的清洁过程。清洁和溶胀诱导参数受淀粉来源以及温度和清洁液氢氧化钠浓度的影响。多变量统计有助于解开清洁实验结果与成分和物理化学性质之间的相互关系和相互依存关系。