Decentralised Finance (DeFi) protocols are usually built by composing a variety of smart contracts. Although composability is a key driver of DeFi innovation, it also introduces security risks: adversaries may exploit interactions between newly deployed contracts and the existing blockchain state to cause economic losses. We introduce a formal security notion for DeFi composability, called MEV non-interference, which requires that interactions with the existing blockchain state do not increase the Maximal Extractable Value (MEV) obtainable from a set of newly deployed contracts. To formalise this notion, we define local MEV, a new measure of economic attacks that quantifies the losses incurred by a given set of victim contracts. We study MEV non-interference under two adversarial models, corresponding to bounded and unbounded adversarial wealth, and establish sufficient conditions and locality principles that enable modular reasoning about secure composability. We apply our framework to representative DeFi compositions including exchanges, automated market makers, options, lending pools, routers, and arbitrage contracts, and show that it distinguishes secure compositions from vulnerable ones. Our results provide a formal foundation for analysing the economic security of DeFi systems built through smart contract composition.
The paper deals with new formulations of a Lagrange interpolant polynomial based on the nodes of the well-known anti-Gauss rule. A first representation is given in terms of the classical Christoffel-Darboux kernel appropriately modified. The second one closely follows the barycentric form of the classical Lagrange polynomial, while the third formulation represents the interpolant as a combination of an orthonormal family of polynomials with respect to the discrete anti-Gauss inner product. Numerical tests show the performance of the explored forms.
We investigate a class of zero-flux chemotaxis-growth systems featuring nonlinear local and non local reaction terms, as well as a gradient-dependent damping, given by {u(t) = Delta u - chi del .(u del upsilon) + au(rho) - b(integral(Omega) u(beta))(delta) - c|del u|(gamma), in Omega x (0, T-max) tau upsilon(t) = Delta upsilon - upsilon + u, in Omega x (0, T-max), where chi, a, b, c > 0, rho, beta, delta , gamma >= 1 and tau is an element of (0,1). Assuming that Omega subset Omega subset of R-n (n >= 1); is a bounded domain with smooth boundary and that the initial data (u(0), tau upsilon(0)) are sufficiently regular, we prove the existence of global (i.e., T-max = infinity), uniformly bounded classical solutions under suitable structural conditions on the exponents defining the reaction term and the dimension n. A central insight of our analysis is that, in contrast to classical scenarios where rho <= beta typically ensures automatic control of the total mass integral(Omega) u, the regime rho > beta makes that the structure of the source term alone does not suffice to guarantee mass boundedness. Accordingly, the global boundedness of solutions is achieved via a two-tiered strategy: first, we identify a parameter regime in which the total mass remains uniformly bounded in time; second, under further constraints on the model parameters, we establish uniform-in-time boundedness of classical solutions in stronger norms, specifically in L-infinity(Omega). These results offer further insight into the interplay between chemotactic aggregation, gradient-driven dissipation, and nonlocal reaction effects, contributing to the analysis of blow-up prevention in structured chemotaxis models.
Internal congestion can substantially affect vented dust-explosion loads, yet aggregate porosity does not describe how obstructed regions are spatially distributed. This study investigates the numerical sensitivity of explosion overpressure to obstacle porosity and spatial arrangement using a FLACS-CFD DustEx framework previously calibrated against an unobstructed experimental reference. Fifteen simulations were conducted for a centrally ignited maize-starch cloud in a vented enclosure, comparing an unobstructed configuration with two continuous rack-like and twelve discrete palletized regions of equal gross obstacle volume. Prescribed porosity ranged from 0 to 0.9; at each matched porosity, both layouts had identical fluid-accessible volume and initial dust mass. The unobstructed case produced a maximum monitored overpressure of 0.341 bar(g). At porosities of 0.4–0.9, maximum overpressures ranged from 0.325 to 0.493 bar(g) for racks and from 0.451 to 0.586 bar(g) for pallets, with pallet values 18.9–38.8% higher at equal porosity. In contrast, for fully impermeable regions, the rack and pallet predictions were 0.383 and 0.252 bar(g), respectively. The rack–pallet ordering was consistent across maximum overpressure and maximum positive pressure-rise rate but reversed between impermeable and partially porous cases. Published experimental benchmarks support the FLACS/PDR methodology for selected obstructed gas-explosion conditions, while dust-explosion studies provide complementary evidence for the combustion modelling framework. The present application, based on idealized rack and pallet volumes with prescribed equal area and volume porosities, is methodologically supported but has not been directly validated against corresponding physical storage configurations. Accordingly, the calculated pressure differences represent internally consistent, model-dependent trends rather than experimentally confirmed severity rankings.
Coastal wetlands are highly vulnerable to biological invasions, yet comprehensive assessments of non-indigenous species (NIS) in Mediterranean transitional ecosystems remain limited. This knowledge gap is particularly relevant for island wetlands, where restricted spatial extent and high ecological value may amplify invasion impacts. Here, we provide the first systematic synthesis of aquatic NIS fauna recorded across 31 coastal wetlands in Sardinia, one of the largest islands in the Mediterranean Sea. Based on literature records spanning 1907–2025, we identified 30 NIS, with Bivalvia, Teleostei and Polychaeta accounting for 73% of all taxa. Temporal analysis revealed a sharp rise in introductions since the 1970s, mainly associated with aquaculture, shipping and recreational fishing pathways, with species originating largely from the Indo-Pacific and North Atlantic regions. Feeding strategies ranged from suspension and deposit feeding to omnivory and predation, corresponding to a broad spectrum of potential ecological impacts. Alterations of chemical, physical or structural ecosystem properties were the most frequently assigned mechanisms of impacts (40%), followed by predation (30%) and competition (10%). Spatial analyses identified the wetland of Cagliari, the Flumendosa River, and wetlands near Olbia as major concentration areas for NIS, with widespread species including Callinectes sapidus, Gambusia holbrooki and Ficopomatus enigmaticus. Using a Comparative Invasion Metric (CIM), integrating species mobility across life stages and local establishment status, we identified marked differences in these assemblage characteristics among wetlands. Despite some detection biases and taxonomic uncertainty, this synthesis provides a robust baseline for assessing invasion patterns and supports improved monitoring and management in Mediterranean coastal wetlands.
