Search

Damage to porous materials in heritage buildings caused by salt mixture crystallization is driven by the surrounding environmental conditions. To understand the crystallization behavior of a mixed salt solution as a function of changing climatic conditions (i.e., relative humidity and temperature), excluding factors such as the internal pore structure, the thermodynamic model ECOS/RUNSALT is the only freeware available that requires simple input and includes the most relevant ions for heritage buildings and solids. We suggest the use of specific terminology and describe how to use the model and how to interpret the output, with emphasis on key limitations for which solutions are provided. When used correctly, the model output can be trusted, specifically when it is used to inform preventive conservation (e.g., environmental conditions in which salt crystallization cycles should not occur). However, salt mixture kinetics and the internal pore structure remain crucial parameters that are not considered in the model. These aspects need further attention to develop a better understanding and correctly model salt damage in relation to climatic changes.

X-ray micro-computed tomography can provide information about the composition and internal structure of materials commonly found in heritage buildings such as natural stone, mortar, brick, concrete and wood. As it is a non-destructive technique, samples can be scanned multiple times, which makes it particularly suitable for studying weathering processes and the efficacy of treatment methods.

JPI Cultural Heritage & JPI Climate, Collaboration between the two Joint Programming Initiatives “Cultural Heritage and Global Change” (JPI CH), and “Connecting Climate Knowledge for Europe” (JPI Climate) began in 2019 and led to the organisation of a joint workshop a year later. Following the recommendations in the workshop report, an expert working group was set up to scope research gaps and opportunities at the interface of cultural heritage and climate change, culminating in the publication of this White Paper. This strategic document is expected to support the two JPIs to generate policy-relevant research outcomes.

Multi-label Text Classification (MLTC) is the task of categorizing documents into one or more topics. Considering the large volumes of data and varying domains of such tasks, fully supervised learning requires manually fully annotated datasets which is costly and time-consuming. In this paper, we propose BERT-Flow-VAE (BFV), a Weakly-Supervised Multi-Label Text Classification (WSMLTC) model that reduces the need for full supervision. This new model (1) produces BERT sentence embeddings and calibrates them using a flow model, (2) generates an initial topic-document matrix by averaging results of a seeded sparse topic model and a textual entailment model which only require surface name of topics and 4-6 seed words per topic, and (3) adopts a VAE framework to reconstruct the embeddings under the guidance of the topic-document matrix. Finally, (4) it uses the means produced by the encoder model in the VAE architecture as predictions for MLTC. Experimental results on 6 multi-label datasets show that BFV can substantially outperform other baseline WSMLTC models in key metrics and achieve approximately 84% performance of a fully-supervised model., Comment: 8 pages, 4 figures

Understanding salt mixtures in the built environment is crucial to evaluate damage phenomena. This contribution presents charge balance calculations applied to a dataset of 11412 samples taken from 338 sites, building materials showing signs of salt deterioration. Each sample includes ion concentrations of Na+, K+, Mg2+, Ca2+, Cl−, NO3−, and SO42− adjusted to reach charge balance for data evaluation. The calculation procedure follows two distinct pathways: i) an equal adjustment of all ions, ii) adjustments to the cations in sequence related to the solubility of the theoretical solids. The procedure applied to the dataset illustrates the quantification of salt mixture compositions and highlights the extent of adjustments applied in relation to the sample mass to aid interpretation. The data analysis allows the identification of theoretical carbonates that could influence the mixture behavior. Applying the charge balance calculations to the dataset validated common ions found in the built environment and the identification of three typical mixture compositions. Additionally, the data can be used as direct input for thermodynamic modeling.