Internally structured porous calcium-silicate-hydrate (CSH) particles were made by a sol-gel process using natural rubber as template. These particles were formulated in cement to produce mortar with excellent thermal insulation properties. The synthesis of CSH and its formulation in cement were comparatively optimized using response surface models based on central composite design (CCD) of experiments and artificial neural networks (ANN). CSH samples produced using a range of rubber/silica mass ratios (0.2–2.0) and hydrothermal processing temperatures (120–210 °C) were used to make CSH–mortars with various CSH/cement mass ratios (0.0–2.0). Thermal insulation capability of the different CSH–mortars was determined. Models based on CCD and ANN both achieved similar accuracy in predicting the insulating response. This notwithstanding, the ANN model identified optimal conditions for making a CSH–mortar that was cheaper and had nearly the same thermal insulation capacity as the optimal product predicted by the CCD-based model. The novel internally structured CSH particles were confirmed to greatly (∼3.4-fold) improve the thermal insulation performance of the CSH–mortars relative to controls.

内部结构的多孔硅酸钙水合物（CSH）颗粒是使用天然橡胶作为模板，通过溶胶-凝胶法制备的。这些颗粒被配制在水泥中，以生产具有出色隔热性能的砂浆。使用基于实验的中央复合设计（CCD）和人工神经网络（ANN）的响应面模型，比较优化了CSH的合成及其在水泥中的配方。使用一系列橡胶/二氧化硅质量比（0.2–2.0）和水热处理温度（120–210°C）生产的CSH样品用于制作具有各种CSH /水泥质量比（0.0–2.0）的CSH-砂浆。确定了不同的CSH-砂浆的隔热能力。基于CCD和ANN的模型在预测绝缘响应方面都达到了相似的精度。尽管如此，ANN模型确定了制造CSH砂浆的最佳条件，该砂浆更便宜且具有与基于CCD的模型所预测的最佳产品几乎相同的绝热能力。相对于对照，新型内部结构化的CSH颗粒被证实可以极大地（约3.4倍）改善CSH-砂浆的隔热性能。